Episode 56- The Africa Puzzle

If you are able and willing to help me with these translations, please PM me on Reddit ASAP. My schedule is getting busier, so any help is appreciated.

DISCLAIMER: I am not a doctor or a healthcare practitioner, nor do I have a master's or a PhD. I am not a native German speaker either- I only know German as a foreign language, one I learned using whatever resources I could get a hand on. If my translation conflicts with that of anyone belonging to any of the groups mentioned above, please refer to theirs instead of mine. Needless to say, if an official translation, done by a scientific translator comes out, then refer to that instead of this one. I am simply a medical student who admires the work of Dr. Drosten and colleagues, and I would like to provide unofficial translations of the podcasts so that they may be helpful to those who do not speak German.For corrections, please reach out to me on Reddit and send advice in. I apologize in advance for any errors I have made. There are some footnotes at the bottom of this translation.

Original release date: September 15, 2020
Original audio: https://mediandr-a.akamaihd.net/progressive/2020/0915/AU-20200915-1644-4100.mp3
Original transcript: https://www.ndr.de/nachrichten/info/coronaskript226.pdf

--------

HENNIG: We have often heard the warnings before the summer. At the moment, the numbers in Germany seem to say that the situation has stabilized since the end of the holiday season. The Robert Koch Institute has said as much much. Where we are now, the beginning of September, we have had 1,000 new cases a day, sometimes fewer.
It seems to be however that for many people, the pandemic has become somewhat invisible. The University of Erfurt is conducting a long-term project in which they ask people what they know about the occurrence of infection, and how they feel about it. It was shown that even though people conform to the measures put in place to control the spread of the virus, acceptance [of these measures] has slightly decreased.
Dr. Drosten, as opposed to early on in the year when the pictures from Italy were everywhere and even in German homes for the aged, old people were becoming sick, the pandemic has become for some people rather unreal. They don't see it soo seriously. Can you see that this is rationally hard to grasp, and that some people have the feeling that everything is actually rather okay?

DROSTEN: Yes, this is the prevention paradox that at least in our country cannot be seen at the moment. Understanding is one thing. Naturally, I'm very occupied with all of this and I often take a look abroad. For me, it is something very visible. That may not be so for others who are occupied with many other things in their day to day lives. Only sometimes do I wonder why people should make such a big fuss. That's a pretty bold thing to say in public at the moment. And sometimes you wonder to yourself—among them, there are people in official positions—if these people really want to inform the public or rather influence politics, and if during the winter, if we have a completely different situation in Germany, would be happy to be quoted on the things they say at the moment.
SUGGESTION: A "TRAFFIC LIGHT" SYSTEM

HENNIG: You are also addressing colleagues from research who have also spoken out. At the very least, if we look at a map of Germany, we can see that everywhere, wherever the school holidays have been over for a while, the figures are at their lowest. Put roughly, the "vacation effect" has decreased. Thus there have been different opinions, as you have said, which say that we must relax things a little. Your colleague Hendrik Streeck for instance has given an interview and was optimistic that adhering to the AHA rules(1) will bring us through the next month relatively safely. He said for instance that a "traffic light" system should be put in place which would show how the situation in every county is like. From this people can adjust their own behavior or the measures put in place. A measure a bit like the rating system in guesthouses with smileys to rate hygiene, you can imagine something like that. Do you think this is sensible?

DROSTEN: First of all I find it as always silly to talk of individuals. I know how it is because people keep talking about me. I would rather have people talking with me instead. Or that people look at what I really say and not always at simplifications from third- or fourth-layer sources. There are completely wrong impressions of what people really say in them. I believe that Hendrik did so at the interview that he had over the weekend. That has also been very strongly simplified. Many have extrapolated a lot from the headlines and the subtitles, and so we have things like, "He supports political change". And he also began the interview with such a statement—and so this impression remains. What he said, rather, in the whole interview, was actually very logical and if many people saw it, they would agree with him. One of the core things that he said for instance, which has been understood completely wrongly after it was paraphrased is that people shouldn't just look at the infection and reported case figures. Nobody does so. This partly because the further you are from being a specialist, the more you rely and hang on to such things. But people who are more "in-the-know" don't do these things at all, they don't look at all at the crude numbers of reported infections. They of course look at additional parameters. But this is certainly going to be paraphrased in his interview, so that he seems to say that people should look primarily at the number of hospital occupancies. Of course, this is a little dangerous. This is taken to be a lagging effect, and thus people should make sure that not too many people go to the hospital because then certain things which nobody can do anything about naturally happen. In the last podcast episode, a lot was said about how at the moment in Germany it is so easy to gauge how things are going within the population and that the numbers cannot always reflect this.

HENNIG: What I just mentioned was the idea of taking a more small-scale look in order to observe things better. What he said, a possible traffic light system, is already being done in Austria. In this county, for instance, you must be careful because you're at Red Level, and in another, you can say to them, "Now you can relax a little". Does this make sense from an epidemiologic and virologic point of view?

DROSTEN: At the moment, we have few reported infections. At the moment, we may be such a situation because we had this lockdown. And because vacations as well made sure that the incidence remained very controlled in the summer. That's good. We could go on like this. Early on in the year, we spoke of similar things in the political sphere. For instance, I can say that in the middle of March, in the Minister President round table discussion, the scientific community, to which I also belong, did not recommend closing schools. Rather, what we recommended is that people look at things regionally. At the time, March 12, the schools were closed in Heinsberg. And we said that this was exactly the right choice of action. Now that we can see that infections are taking place, we can do something like this, even without knowing how exactly schools contribute to this, and close the schools as a precaution. Thus we went out of that meeting and politicians implemented a nationwide school closure, because one senate chancellor after another, or chancellery in each federal state, decided on the same thing. Then there was a school closure throughout all of Germany, which was in the end attributed to the scientists, which is not true. People can think about such things now. However, back then it was very plausible that the incidence of infection was completely locally distributed. Back then, the infections were only just coming to the fore and local amplification events were appearing. Now, things are different. Now we have only a few cases being reported. But we have a situation where many young people are being infected. Their symptoms are not so obvious. Maybe their inclined not to allow themselves to be diagnosed immediately. This leads to outbreaks which catch us by surprised. Thus, we should look very closely at this, which we are doing. We are testing many people, but it is still difficult to do this locally. Right now, we should say that there are very few infections, that is perhaps the general impression, and it is certainly true for Germany at the moment. But there's no guarantee that things will remain this way for very long. Right now we're looking at individual counties in southern Germany. And we must also look at neighboring countries.


MEASURES WITHIN SCHOOLS
HENNIG: In Germany, we also have a few outbreaks in schools. Regarding the topic of schools, as you have mentioned it, for instance school closures, there have been arguments here and there supporting a harder approach to non-pharmaceutical interventions. There are opinions from researchers and even parents that are pushing for requirements to wear a mask in schools, here in Hamburg for instance. The thought behind this is that with stronger measures, the numbers can be decreased to such an extent that we would be able to create good conditions for the winter. Could you respond to these arguments?

DROSTEN: I am not sure whether the general wearing of masks in schools would be a tightening of measures so strong as to let us say that we really are lowering the incidence rate. Wearing masks in lessons certainly does decrease the risk of an outbreak within the class. That is, however, to do with class outbreaks, and we should pay attention to that. There are other arguments as well, which are more fundamental in nature, which say that efforts should be made towards very strong, wide-reaching measures in order to win more time during the winter. From a purely physical or epidemiological perspective that is, we can say that this is completely the right thing to do, but from a societal perspective it simply isn't tolerable. In Germany, we are in a good starting situation and must take advantage of this moving forward. What is more important is, first of all, that we do not become complacent, and second, that we do not send out messages in public that are complely counterproductive. The general population must vigorously keep alert and not do the opposite. To say, "Ah, what we did back then was not all completely necessary. With what we know today, we would have gone about things completely differently back then." —we should certainly not have that. For instance, when things are done while looking at only one branch of science.You can't just summarize all of this in one sentence. From this, a perception may arise, one which projects towards the future, in which someone would very quickly generalize something like this, make broad statements and say, "Well, we achieved well over our target early on in the year, we really don't need to do much more. You can see that even after all this time, we don't have any new cases of infection." These are general considerations that are very much in the fore at the moment, and really, you must really think twice, thrice, about how you behave regarding these things. And if you're standing before the public, you must really also think twice, thrice about how you want to speak about these things and, even set things in stone, write them down, in case you will be quoted on them later on.

DOUBT AND CRITICISM
HENNIG: I would like to go through a few arguments step by step. There is the question of what these measurements have actually done. Doubt is always spoken about as if it was evidence, namely about the effectiveness of the measures. Can you really, very generally speaking, judge the burden of these? There are models for this. There are of course no real control groups which have gone through the pandemic without measures. Thus, it is always the case that you can't really find this out, no matter how good the models are. Most of all, you can't judge each one individually and separate them from each other.

DROSTEN:We have already did a whole podcast episode on that topic. The article mentioned then continues to be valid. Even in Germany, this discussion has gone on on a completely different, much lower level. A paper has appeared on the website "Evidenzbasierte Medizin" [Evidence Based Medicine]. There, arguments which I thought were already discussed in Germany at the end of the early part of the year, were repeated, about how the Rt value had already gone below one before the general contact limitation measures, the lockdown, was implemented in the middle of March.

HENNIG: The reproduction number, we must return to that.

DROSTEN: Exactly, the Rt. This has really already been discussed, that this contradicts data that shows that the mobility within the population already decreased in the first half of March, before large gatherings were stopped. And that the stoppage of large gatherings came before the decrease in Rt and that therefore Rt isn't everything. Mathematically, this isn't completely true, but to put it simply, Rt is only the first derivative of what happened, it is the trend. But if you want to lower something, the trend should first be reversed. That's exactly the point. So this entire argument is really neglect of scientific principles. That this opinion piece came from something entitled "Evidence Based Medicine" is something that I have been quite taken aback by.

HENNIG: An argument from this paper that has been repeated again and again is that the "infection fatality rate" is actually low, in other words the proportion of deaths to the number of people who are infected, and not just those who are ill. Now we can actually see that the infections have risen slightly since the summer. The number of deaths however does not follow this trend. You can see that, even in hospitals, there haven't been any dramatic situations. Does this relation support the argument that the people who really are in danger are actually the old very old and that young people don't have to worry as much? So the argument goes.

DROSTEN: This is really a society-wide problem. You can't completely separate the old and the young. Another example: this paper makes arguments over the mortality rate among the infected looking back over the last four weeks. This is itself a contradiction. You can't talk about such a rate over such a period of time. Nobody at all can have this data. And in this paper, a lot of evidence that has come to light has simply been ignored. There are two very, very good studies based on serological studies on representative populations and on good, officially reported data, in Spain and in England. These are two countries in which we had a regular first wave. Here we have death rates among the infected which are based on large, population-representative studies made specifically for this purpose, and not on data which has been distorted by confounding factors. In England, we have a death rate among the infected of 0.9% and in Span, 0.83%. There is little to argue about here. In a population of this age distribution and morbidity and this burden of disease and so on, this is something like a constant of nature with this virus. In Germany, we are structured similarly to Spain and England. And one should not argue that in Germany, in the last four weeks, only a few people have died. That is trivial, that is clear, that's just as well. We have somewhat controlled our epidemic. This success is due to the combined efforts of science, medicine and politics. One can say that. And people should not behave as if this is only a hallucination, and dare to refuse to look at the countries directly bordering us. That's quite foolhardy.

HENNIG: Another point of criticism, which I have spoken about last week in the previous episode with Dr. Sandra Ciesek, because we had spoken about the theme of symptomatic vs presymptomatic and presymptomatic infections in depth, is the challenge to bring more transparency to the statistics, to differentiate between symptomatic cases, to show that this group is made up of the truly sick and these are [just] infected. Can this be done?

DROSTEN: As a matter of principle, the Robert Koch Institute does this already, this separating into groups, into age cohorts and so on, the Institute doesn't do this too badly. This would take longer to do in other countries. And it would be nice for this to be even more exact. I ask myself whether these more exact numbers would make these rumormongers in the public argue over something else or stop arguing so destructively. After all, that's what it is end the end, people try to do something here with combined efforts which other countries have not done so well. And our politics does well in this regard. Not everything is perfect, but the end result is still good. And when you look at balancing this with the economy, things aren't too bad. Especially when you remember that Germany is an export economy, and so is for the most part beyond our control. But our economic damage is still damage to an export economy, and things are doing poorly in other countries, which we cannot control. Whatever we can control here in Germany is doing well. And still there are these destructive claims which suggest that not all of this was necessary—that is as intelligent as to say during this beautiful week late in the summer that it never rains at all. "What are we making so much of a fuss about in the autumn? Why are we talking about clouds and rain and this weather? Let's have a look outside. Everything is great and the last weeks were all completely good in terms of the weather."

HENNIG: As we have spoken about "Evidenzbasierte Medizin" already as well as its paper, I would like to bring it back up again. It speaks very explicitly about 'an indiscriminate over-diagnosing", because testing is not limited to people with symptoms. Would it make sense for only high-risk groups to be tested as a matter of principle, namely those who have had contact with an infected person or have symptoms themselves?

DROSTEN: Yes, this term 'high-risk groups' is really used in this paper and is completely inappropriate for this context. I don't want to spend any more time at all on this paper right now. There are many errors inside, even in the use of scientific citations, in the reading of the respective articles. For instance, a certain figure was given for how much a "quality-adjusted life" throughout lockdown conditions would cost. Then they argue about how much this costs. That the original article did not argue about this at all but rather said, "How much would it cost if the lockdown was prolonged over the entirety of June?", was completely overlooked. In other words, here we have a paper about evidence based medicine, where the literature, in other words, the evidence cited—this is the principle behind evidence based medicine—was obviously not read at all. I've asked myself why this article was published without naming any authors. On the contrary, with opinion papers from commissions and specialized institutions, we have a list of authors, because the people who write them must be responsible for them. I've wondered about this [the lack of authors]. And now to the many things that have been said about testing. I believe that even if we ignore this opinion paper, a discussion over laboratory testing is being developed. In Germany, we are well enough equipped financially to have been able to ensure that a lot of testing was done over the summer. We have already said that testing returnees from vacation is only reasonable within limits due to epidemiological and financial reasons. It pushed laboratories to their limits. Politicians have already taken measures long before. It's a matter of principal. And at some point we should say, "Okay, we've had enough of this topic. We've talked about it. Sometimes things went too far, it doesn't matter, in the end a good discourse has taken place and politicians have long since reacted to it. Now these accusations must stop, it's time to look to the future." This means that within the realm of PCR, of medicine, of targeted diagnostics, from a patient-centered perspectives, testing should be symptom based. And of course more should be done regarding the antigen tests. We've mentioned this already. They look good in terms of validations. Now there are regulatory concerns. A good compromise must be found between conformity to the law and applicability, so that these tests come to where they are needed. For instance at the entrances of senior housing facilities, where they can do an unbelievable amount of good. We would then suddenly have the possibility to confront situations that may be too much to bear in different ways, for instance, if the elderly can no longer be visited, and proceed from there. That's only one example. But we should also note that we should not use poor quality tests.We have a responsibility to ensure this. This is very clear for instance in the example of the senior homes. If we were to use a flawed test, one that occasionally fails to detect an infection, then we would have an outbreak in a home for the aged with a corresponding death toll some weeks later. Thus, we should have a good compromise between reliability, conformity to regulations and speed. Of course, politics is productive in this regard, and so is industry.

WHAT SHOULD WE LOOK FOR IN TESTS?
HENNIG: False negatives—such is the case of the tests at the nursing home's door—have massive consequences. The question of false positives is of course not completely unimportant for the question of the acceptance of the tests. How can this problem of pretest probability be communicated? That is a statistical problem. We've also mentioned this too before in our podcast. Theoretically, if the prevalence is low, in other words if the virus is present less often in the researched population group, then the false positive rate increases. Must the test strategy be adapted depending on how the infection behaves?

DROSTEN: Yes, certainly. We must not get into testing theory here. That's what some parts of society are doing, having lively discussions about testing theory but without any medical practice. And they then say, "There are some specificity figures and we're calculating that on top of the tests." And then we say: "All of the detected in Germany, they can't all be true positives, they're all false positives"—suc nonsense. If it was so easy, you really wouldn't need to study medicine at all. Then anyone could just open a laboratory.

HENNIG: But there are false positives when there are few infections—that should be said.

DROSTEN: There is certainly this effect, exactly. But if we detect a positive in the laboratory, we look further. There are results that are so clearly positive—they have a quantitative aspect—that we don't need to think about them any more. There are results that are barely positive, which should be tested further. Then the samples will be tested again and so on. People don't see what is going on in the laboratory at all from the outside. But even at the lowest incidence rates, we never have a situation where false positives appear one after another in the statistics, of which none at all are infectious cases. The laboratories don't work like that.

HENNIG: Viral load also plays a role.

DROSTEN: Yes, that's one of the criteria. But then additional tests and so on are done. As for your question: of course, one would measure other things as well depending on the current epidemiological situation, in other words how common the sickness is at the moment. For instance, during the middle of a wave of infections in the winter, one would be happy to have antigen tests. And even if they know that sometimes they end up showing false positives, that wouldn't come into consideration at all. On the other hand, many cases would be overlooked at the end of the spectrum, with sensitivity, because these tests aren't perfect. But they have a huge advantage: they are quick and can be used anywhere. And that is decisive. The advantage of speed with a rapid tests outweighs by a large margin the advantage of PCR diagnostics, sensitivity at the cost of days because of logistics. What use is a very sensitive PCR whose results I have to wait three or four days four because the laboratories are overloaded? Thus, during a possible winter wave, which we hopefully won't have yet we still have to calculate presently, we won't be able to go about things with mass PCR testing. We will really need these antigen tests. The concrete things which we must argue about are not textbook subjects that aren't completely understood such as testing theory, sensitivity, specificity and predictive value, but rather, what we must discuss about as a society, at the very least politically, is regulation. We will not be able to do this, to validate these tests as "home tests", before an ostensible winter wave will be over. The amount of effort required is too high. For instance, results will have to be provided, showing, to put it bluntly, that all of this is idiot-proof. That isn't to do with the parameters of how much effort the tests require, but rather, the firms must show that the average person won't get everything wrong when applying these tests. For this, observations of how they are used are really needed. And this cannot be done in time. This means that a compromise is to be found when we say, "This is a technical, published test which has been released to medical specialists." And now we must consider how we define "medical specialist", by what criteria? Let's use high society going to plays as an example. Say a large, famous theater organizes a play that has been planned for a long time and has to be shown. And then people will ask: can we do these rapid tests at the ticket booth? Would that be possible? These are things that are already being discussed in public. And to that I say, and I say this despite not having given a lot of thought to it...maybe lawyers would say, "What Drosten has said in his podcast is again complete garbage." Maybe that is true, but I will say this in spite of the danger, simply as an example: can this theater bring in a medical-technical assistant for the duration of the play, so that they can test? Or must the organizers bring in a laboratory doctor? Or would it be enough for the booth cashier to do a two day crash course? And would they then be qualified enough to apply these tests? These are questions that we must really discuss during the coming weeks. Maybe not unrestrainedly in the general public, but in politics, in the backrooms of the ministry departments, these things should be thought about. Not just within the ministry of health.

HENNIG: These are questions that, if cleared up, would possibly offer a perspective for instance to the culture sector, which have had to avoid everything for a very long time.

DROSTEN: I only mentioned this as an example. That goes for many sectors, where this would be useful.

THE CORONA PANDEMIC IN AFRICA

HENNIG: We have also prepared for today another topic of discussion. Up till now, the questions were about the relationship between test numbers and the infected. This is a pandemic, of course. For this reason we will take a look at what is happening in general, globally. Some of our listeners have asked us to mention the situation in Africa here in our podcast. We want to do this today as well. This seems like a sweeping generalization, Africa. We in Europe tend to talk about "Africa" all the time. There are over 50 different states there, very different realities and often heterogeneous situations. But there's something noticeable there—looking generally, maybe with an exception that we'll talk about later, the dramatic progress [of the pandemic] that many African countries feared seems to have not occurred, is this true?

DROSTEN: Yes, it seems that way, up till now this has not happened. At the very least, things that were forecasted early on in the year when certain models were applied, and people didn't know that there was such a strong imbalance between age groups in mortality. Namely, that the elderly would be so strongly represented in the death statistics and, given the population and so on, it was calculated that in large African cities, we would see such problems as those early on in the year, for instance in northern Italy, would be seen, where—as we remember—the dead could not even be transported normally, but military vehicles had to be used. If you make projections about a large African city, where these sorts of logistics aren't present, then you would worry quite a lot. In the large African cities, we seemingly have not seen similar pictures as of yet, at least not in official releases. I must say—I cannot at this moment explain this situation in Africa in detail. I must say that this is a problem that really bothers me. About two weeks ago, I remember, a study about Kenya was reported by the BBC, which has appeared as a preprint. This is a study based on laboratory data collected in Kenya and on an initial observation that in PCR studies, the PCR detection rate has been sinking since July. Serological testing and mobility data was also used. And in the end, by modelling these underlying data, the conclusion was that despite a rise in mobility in Nairobi and Mombasa, two large cities in Kenya, where these studies primarily took place—there was a strict lockdown in these areas—there was no rise in infections. And the conclusion drawn from serological data is that in these areas, herd immunity has been reached. That blew me away. I'm not sure if this will continue to be the case. This is a publication that has not yet been peer reviewed. But we probably should talk about it a little.

HENNIG: Maybe we should ask how significant these infected figures are first, even if this can't be answered, before we look at the question of antibodies, because then it will at least be reported that far fewer people were tested in many African countries. I have a statistic here: in Great Britain, for instance, there are 200,000 tests per 1 million residents. In Tanzania for instance, which doesn't report any more figures at all, there are 63 tests per million residents. As for detected infections throughout all of Africa—according to the African Centers for Disease Control, there are 1.3 million detected infections out of 1.2 billion residents. Are such figures significant, knowing that there are supply bottlenecks with test kits, that there are armed conflicts, all of which prevent large-scale testing?

DROSTEN: It is impossible to make statements about the whole of Africa. I believe that we must stick to a few concrete examples. Unfortunately, this week we must stick to scientific literature that has taken a long time to come here. In African states, there are reporting systems of very different quality. Overall, it is clear that the reporting systems there aren't so efficient. You must imagine that we have a huge spectrum [of the quality] of infrastructure in African countries. For instance, if in a country like Ghana you drive from Accra, the capital city, to Tamale in the north, it's a bit like traveling back in time. You can't expect there to be equally reliable reporting throughout the entire country. And Ghana is a very well developed African country in sub-Saharan Africa. There are of course completely different situations. An example that I can name, and which I can support with literature, is this study, which has appeared in a preprint, about Kenya. Kenya is one of the countries that is well known for having built very good medical infrastructure over a long period of time. Since the beginning of the epidemic—this was happening even in March—320,000 PCR tests were done by August 10. That is certainly one of the highest numbers of PCRs that was done in Sub-Saharan African in one state. However, that is comparable to the number of tests that we alone had done in one week at the end of March. And then more were done. This is the PCR test basis for the entire epidemic up until August 10, which has been factored into this public statement. Within this time there were 24,000 positive PCRs. This is about seven, eight percent—a lot. This large fraction of positives is due to the fact that there was little testing done, that testing was done in a population with a lot of incidences. The question is simply: how punctual is this testing? This is not always clear from such studies. And how exact are the other numbers that are calculated from this testing? For instance, the number of deaths reported is dependent on testing. In a country like Kenya we can't say that anyone who has died from a febrile illness, perhaps with some respiratory symptoms or not, is a case of Covid-19, of SARS-2. This must be confirmed in the laboratory. And on the one hand there is the number of confirmed deaths, which is 391 throughout all of Kenya during the evaluated time period. You must think about this a little. In many countries, this is not related at all to the higher-than-normal death rates. And during the peak of the pandemic, for instance in European countries, we could ascribe many of the higher-than-normal deaths to the virus. But in Africa, it is relatively unclear what one can take from this, especially in countries in which few people were diagnosed.

CAN PEOPLE TALK ABOUT HERD IMMUNITY?
HENNIG: You've mentioned this figure for Kenya, for Nairobi, in which seven to eight percent [were infected] out of all the PCR tests, but have presumably tested the areas with many infections. When people hear the phrase "herd immunity", that's a whole different matter. That means to say that the numbers are so high, there are so many weak or completely asymptomatic infections, that the infection has already taken its course within the population. Only antibody tests can provide conclusions about this.

DROSTEN: Exactly, this has also been done, it is a crucial part of this study. Here, this is really a little hard to understand. Here it is said that 3,000 blood donors were tested and their seroprevalence was around 9% at the end of the study period.

HENNIG: 9% show antibodies.

DROSTEN: Correct. And that is of course an extremely high number for blood donors. Blood donors are healthy people who go off to give blood. And this was shown in a study that was conducted on the same cohort in May. Back then, the seroprevalence was 5%. I'm looking at this from a laboratory perspective. By the way, I find it increasingly important to note that in Germany, we have something of a weakness in infection epidemiology, this is for us a weak specialty. Microbiology and virology are in comparison stronger specialties, the laboratory-based specialties. In other countries, the strength comparisons are a little different. However, I also have the feeling that a very critical look at laboratory data is missing from some epidemiological studies. This direct transfer of lab data into epidemiologic modelling can be problematic. I believe that such a problem is present here in this study.

HENNIG: Because real life factors of uncertainty are not included?

DROSTEN: From real-life laboratory life. For instance, there's a reference to some research which explained how this serological study on blood donors was done, in other words, a previous, published study. This concerned the state of affairs in May, in which antibodies were found in five percent. If you look at the tests used, there was a single ELISA test. That wasn't a commercial ELISA test, but rather an in-house test. This means that a protocol from an academic work group was then apparently imitated in an African laboratory. And thus you have the figure, five percent of blood donors. However, if you have experience with serological studies, then you'd know two things. The first is that with serological studies, it is difficult to take and imitate the protocols. People want to buy industrial products that were made according to quality standards, with no weaknesses in test quality. That's one thing. The other is that when you use such industrial products for serological testing, then you often come across a really unpleasant surprise when you test populations that were not taken into account within the framework of validation. Simply put, such serological tests must be modified and then set up. You conduct studies that test completely healthy people first, those who don't have the illness. In this way you ascertain the "background noise" of the tests. And then you say, "The threshold at which we will call the test positive is to be placed quite above this uncertainty level." However, if we move on to different populations—this is sadly a notorious problem in African populations because different infections are constantly present in the population which we do not have...If we go to Africa with a European-validated serological test which has not been validated for African populations and whose thresholds at which we declare a positive tests were not readjusted, we would then often see many, many positive results which aren't actually true.
For instance, Felix Drexler made a commercial serological test for the SARS-2-virus with me at the Institute. He used it in Africa. He saw up to 20% false positive results. This is a study that has already been peer-reviewed and accepted for publication. We could include it in the log of the podcast. That was only a very small study, a technically exemplary study. I'm only using this as an example to remind people that there are so many pitfalls with such seroprevalence studies and that people must be very careful of putting such data in epidemiological modelling studies which in the end lead to such important conclusions such as, "Herd immunity has been reached in this country." People must be completely clear about the political implications of this. At the moment the EU in particular, but also other countries worldwide, are working together to ensure that poor countries, countries of the global South, are supplied with vaccines. In other words, that countries do not choose to go about the acquisition of vaccines alone, with supply contracts with industries, but rather say, "If somebody secures a supply of vaccines, then they should also pay for poor countries through a common mechanism." It is of course not politically beneficial for scientific studies to appear which say, "But why? The problem in Africa has long since been solved, the plague has already ran its course. And all is going well, hardly anyone has died." These are very difficult perceptions to make now. First off: is this really so? This work predicts that 40% of people in Mombasa and Nairobi approximately—let me flip through again—around 40.9% and 33.8% of people in Mombasa and Nairobi show seroprevalence or have been exposed. But this is based on relatively unclear and error-prone serology data. And then there are these future projections and models underlying everything. If this was really the case, if the population of a large city like Nairobi continued to be cautious and where people couldn't move around completely freely and other cautionary measures were in place was 40% immune, I would approve of this. Then I would say, "I believe that a level of immunity is present in the population at which we can talk about herd immunity." This isn't 70%. But it must be said that the authors have taken this argument very well in this study. I find that this isn't a bad study. I only find the shortened conclusion of this somewhat problematic. The authors give this argument that maybe herd immunity has already been reached because people aren't mixing freely—see the previous podcast episode in which we spoke about this(2), that not all transmission networks are available at all times. Not all of the population is susceptible to the virus here and now, making the real herd immunity threshold naturally lower than 70%, over a year or so. Over a whole year, we need 70%. But right here and now, maybe around 40% would be enough to stop the spread.

HENNIG: Because I continue to move around within my circles.

DROSTEN: Correct. And there are also other circles that you're concerned with. I believe this is so. And I'm happy to think that at the moment the observations say that the rate of PCR detections is lower. But whether this is because herd immunity has been reached is something I am not sure about, whether the studies bring enough evidence for this. This is always the danger: a study is done and then goes through the BBC and is generalized. And then people say immediately that in Africa, the whole thing's over with.

HENNIG: However, there is another antibody study from a region in Nigeria that has yielded similar results. In there, blood donors weren't researched, but rather health system workers, among others.

DROSTEN: There are many studies, studies which are as of now preprints. And this should be said: be careful, be careful, they have not yet been peer reviewed. As a virologist, I would like to say a few more words of caution. This study in Nigeria for instance has also taken place in a state, Niger(3), in which there was an EGG seroprevalence of 25.4% between May and the end of June. However, there are things that should be noted with this, first off, only a small number of participants were investigated. Only 185 people were investigated in total. One must ask themselves if such an extremely populous country such as Nigeria can be represented by 185 participants. The selection bias, the bias affecting how participants are selected, is particularly big here. It may be that in the area investigated, an outbreak was present and thus the study was done there. Or perhaps someone said, "Dear people, today we're doing a scientific study and whoever wants to can participate." Then of course people who were just sick would come because they would want to know whether they had this sickness.

HENNIG: This has been commented on, there were people with symptoms who said, "We had flu-like symptoms in the past."

DROSTEN: Exactly, this is of course a distortion of reality. Then you have the fact that these 185 people are divided into people who live in the city and people who live in the countryside. And between them, there are no differences to be seen. This is very strange, that the same prevalence was to be seen in the countryside as in the city, during a very short period of time, from March until May, when the study was done. How can that be? The differences between the cities and the countryside in Africa are extremely big. From my own experiences, when you get in a car from the airport and then to the hinterland, it's like you're travelling through time. It's an impression you get when you're in Africa for the first time, one that doesn't leave you for months. You can't experience anything like this in Europe. And I find that, for such a disease of this type to have gone through the population in this manner, is something that is at the very least in need of a second look. And if you take a close look at what was done, a serological test using the lateral flow process was done. In other words a test that we have already spoken about in the past, following the principle of a pregnancy test with antibodies. And we know that these tests have their flaws and that they have not been validated for African populations. And in this study, what must be done in these sorts of studies was not done, namely, that before the pandemic, you take an equally large number of people from the same country, where the virus had not yet been detected, and see if the test is negative, as would be expected. This has not been done. In the study by Felix Drexler at the institute for instance, this was done. And before the pandemic, you could see very high rates of positives—which cannot be true. The virus was simply not there at the time.
We can take a look at another study which comes from Malawi. In terms of quality, the test was done much better. Medical personnel were investigated and 12.3% were found to be positive for antibodies. Here in the study in Nigeria, by the way—I don't know if I've already mentioned this—there were 25%, in other words really a lot. Here in Malawi, 12.3, but it should also be said that this was done in a hospital. These are medical personnel who are involved with patients. The investigation was done in a city of 800,000 residents. And it must be said that this city has a long colonial history, with many vacationers, in which many people from outside the country come in and go on vacation, with a lot of tourist traffic and in which therefore the virus naturally slips in through different methods. And this is really something that should be made clear for Sub-Saharan African countries in general: the cities are very different from the countryside. So we have the phenomena of rural emigration, which leads to us having many young people in the city, against whose numbers the elderly pale in comparison to. And this is precisely the impression, the side of the coin, which we have not spoken about yet—the impression that the death rate, the number of deaths, is too few in African population. That can be because at the moment, practically the only news that reaches the outside is news from large cities, and that medical studies are practically only done in large cities. And reports primarily only come from large cities, while in these large cities, we have an age distribution that is not at all typical for the entire country. The city is simply much, much younger. We in Germany are not familiar with such phenomena, that in terms of city demographics we have very different compositions between the city and the countryside. This leads of course to the fact that the virus affects the city much less. And we don't know at all what is happening in the hinterlands. Whoever has been to Africa knows how many old people there are in the villages there. There, the people are very old and a part of them have health problems. And we probably don't have any reported figures at all from the countryside in many Sub-Saharan African countries—never. So we will only be able to see the highest estimates of deaths above the average in retrospect.

HENNIG: And in the countryside it is often a problem, protecting yourself with hygiene measures. That buzzword, "clean water", is of course something that should be taken into account.

DROSTEN: Many things, many things. Another thing is perhaps immunomodulatory things, helminth infections, which are extremely widespread in such populations. [Populations] where we don't know at all how this affects the pathogenesis of this disease. So we have here immunological pathogenesis. And we don't know what such a modified immune system does to the lungs during the course of this infection as opposed to the immune system of a European.

HENNIG: Because parasites may subdue an immune reaction in order to remain within the hose and then cause heavy inflammation again.

DROSTEN: Exactly, because perhaps the self/non-self recognition in cellular immune systems are a little more loose, to put it colloquially. Immunologists would probably smile about this [analogy] or get upset about this, depending on their disposition. I believe most would smile and know what I mean by this. With immune pathogenesis, we are just starting to understand what happens in the lungs. Of course, an infection that is present throughout the population, which infects everyone from the youngest ages onwards, means that completely new adjustments must be made in order to understand this. You can't simply apply findings from Europe to Africa.

HENNIG: This means, however, that certain conditions that sound dangerous at first—the immune system may be affected by completely different problems, for instance—can also have positive effects regarding the virus.

DROSTEN: Absolutely.

THE EXAMPLE OF SOUTH AFRICA
HENNIG: There's still the question—I've just looked at it—yesterday, about 32,000 deaths were reported in connection with the coronavirus throughout all of Africa. It doesn't matter how far this question of possible herd immunity may be interpreted, the big question for virologists is: there clearly are few serious courses of illness. An explanation may be, as you have said, that the elderly make up a smaller proportion of large cities. What other factors can influence this?

DROSTEN: Honestly, the main factor that I see at the moment is reporting. I have my doubts about what is at the moment being transmitted through big, widespread announcements. I will give an example of why I'm so much in doubt. We know that in South Africa we have a very, very good public health system. Relatively many people were tested, not on a level comparable to European countries, but enough. And data's already starting to come out. And as for myself, if I want to know something, then I can start to contact people. A really good acquaintance of mine is a virologist in South Africa. He's even given me better data at times. This is public available, but you won't always find it if you just Google it from Germany. I got data from the public health system which is also known and discussed about in South Africa. For instance there's a district which is very well researched, the metropolitan region of Cape Town. We know a little in terms of numbers. And you must imagine that Cape Town is, at its center, built in a very European fashion. But in Khayelitsha, one of the large townships in South Africa, we have a very, very poor population. We have high HIV prevalence rates, all these problems that South Africa also has. There, we have good numbers. One can make a very interesting observation, which I also believe, when such studies are done—that 40% seropositivity is to be found there. This is based on studies of blood tests from the field of pregnancy pre-examination. That's a good look at an adult population, pregnant women.This age group, pregnant women, is younger than pregnant women here, mostly people in their 20s. And there are the HIV clinics that are very widespread, which have been created by the public healthcare system. There, there is a high prevalence of HIV and therefore a lot of treatment and patients, which you see again and again all the time. And if you look at the blood tests parts of the population for SARS-2 antibodies, you would find 40% seroprevalence. This is a sample from the poor population. Public hospitals were used here. Therefore, poorer people tend ot go there: the rich tend to get themselves treated in private clinics and in private practice. With 40% seropravelence, we can observe that the spontaneous spread of the virus happens less often. Fewer deaths will be observed and fewer new reports. Over the summer, in the present winter—it is in the southern hemisphere after all, subtropical climate, so not a cold winter, but still winter, perhaps like in Italy—then, the virus spread very strongly, especially within poor populations. There really were problems in medical care. That did not make an impression like Kenya for instance, but it does make an impression—here, we have a big pandemic problem. Regarding the situation in the metropolitan area of Cape Town, there were 3,900 deaths more than average, "excess deaths", and 40% seroprevalence in the 3.7 million-large population in this city. If we can say that there was, perhaps this is a bit large, 40% seropositive, then we could say that 1.48 million people were infected during this wave in Cape Town. There are 3,900 deaths more than average among them. If we divide this, we get a value of 0.28% for death rate among the infected. This is a realistic number, even for a European population. Now we're within the confidence interval of the estimate. I said before that in the first wave in Spain and England, we had 0.8 and 0.9%. Now we have 0.25%. You should look at this a bit generously. I of course calculated a seroprevalence of 40% primarily in poor populations. This probably won't apply to all of Cape Town. Maybe the real seroprevalence is much smaller. The African scientists in South Africa also say as much, that the real seroprevalence is  probably smaller and may be half of that. If so, we'd be around the area of 0.6% [infected mortality rate]. And then we would be right within European estimates. I can't recognize at the moment why African populations should be less affected in terms of deaths, especially when we realize that the worst point of the pandemic affected areas with poorer populations. That is why I find it hard to believe just like that, that Africa is not so affected by all these problems.

HENNIG: I would like to briefly ask about South Africa. The gist of the data you presented to us is that we simply know more about South Africa and thus the situation seems a bit more dramatic than in other countries because we have more numbers and more valid findings?

DROSTEN: First of all I believe and hope that South Africa has, after the difficult first wave, really gone the way of herd immunity and especially within poor populations, who aren't very well supplied, taken steps towards population protection, so that one can calculate if that the continued precautionary contact measures are also especially effective. The very simple consideration of what mask you wear, which may not be completely effective, may be actually especially efficient there because in addition, there's also population immunity. That would be nice, it should be said, that was very nice to hear. However, that should not be taken to mean that regarding international cooperation and aid to poor areas of the Earth, less care should be taken of the African continent. On the contrary, people must really pay attention to what is going on there because we know far too little and we may experience nasty surprises if we let ourselves grow lax because of these studies, which appear and may be over-interpreted. And back to the estimates, if you look at the Johns Hopkins figures for South Africa: there are 15,447 registered deaths among almost 650,000 registered cases, a 2.4% case fatality rate. That's a case fatality rate which we also see in many other countries with respect to the reported cases. That's not the infection death rate. These crude statistical data is always estimated too highly. But they are not estimated differently than in European countries. Thus, from my far-removed perspective, it is a little careless to say that things are completely different in Africa. Certainly, in large African cities with an especially young age distribution, yes, that may be, you may see fewer deaths. But it isn't just population affecting the deaths.

HENNIG: There are still many open questions, however, about how Africa is doing. To close, I would like to give another question from the listeners. A cardiologist asked this of us. This is somewhat related. Immunologically, what really can happen, and what has happened? Is it really plausible, he asks, that the immune system—you can also ask this question of Europe—that the immune system is trained when one often comes into contact with a low viral load, in other words encounters infected people often, who may however not emit much virus or if you don't get much virus—to speak casually—in other words, a habitation effect that arises with the Coronavirus and then, when you encounter a very highly infected person, a weaker infection occurs? Is this plausible?

DROSTEN: Yes, it is plausible. In the New England Journal of Medicine there was such an article recently about this perspective. This was really just an opinion piece. But this has gone very far in the media and regarded as new scientific knowledge, which it isn't. This was the argument. If everyone wears masks, then everyone emits less virus on average. These lower viral levels is reminiscent of a measure used in pox vaccination so-called variolation, in which scabs from pox victims were taken and dried. Thus, there was noticeably less virus in these scabs, but it wasn't gone entirely. And then these scales were scratched into the skin in a sort of earlier procedural form of a vaccine, with real, unweakened virus. But because there was so little virus, many people were thus not fully sickened and didn't die. But at the end they were only mildly sickened and then immunized. And thus this consideration was made: Does this exposure to the virus while wearing masks, if everyone wears masks—is it true that the average transmitted viral dose is a little lower and then it isn't the full infection which occurs, but rather a superficial infection and a slow burnout? That's a bit of an old term.

HENNIG: A good one.

DROSTEN: Without us noticing it, we are suddenly immune, because we have only received small amounts of the virus. There are reasons to think that something like this can happen. For instance, in a previous episode we mentioned this Swiss study on seroprevalence, in which there was evidence that hospital personnel, who were exposed as part of their jobs, showed evidence of light secretion of IgA antibodies on mucus membranes, even though there were no antibodies to be seen in the blood. It may be, then, that they are partially protected. And sometimes, experts in the public use this term, "partial immunity", which has up till now not been heard. This belongs, rather, to the same area as malaria immunity. We know that partial immunity exists. However, it has nothing to do with respiratory tract infections. As well as with variolation, the pox, there is a completely different mechanism of infection. You breathe the virus in, but then it spreads systemically through the blood. I would not want to dismiss all of this. This may be true. And it would be great if it was.

HENNIG: But there's still a lot of conjecture.

DROSTEN: Yes, but this is all great academic speculation. All of this is interesting. But the question is who would want to take responsibility for implementing that into official guidelines? Nobody. And that's really the problem we have in the autumn and winter. In the end, we haven't been holding an opinion podcast for academically interested friends of virology for a long time. It was good while it lasted. But this is always being asked—and if we don't speak about it here, then others will from the outside—what doe all of this mean? And this will quickly generate a message in this direction, in which we started to speak about today—it was all in vain. One could've known this and only required mask-wearing and so on. There are so many possible objections. We're getting into the hundresds, into the thousands of arguments. But if you want to argue back, you must say: What is the point of throwing all these accusations in retrospect, which are all essentially wrong and which look forward into the winter, even perhaps regretted by those who now make these reproaches? You really only have to look at neighboring countries. And then the argument will come that yes, but nobody at all dies. Nobody is dying now, of course, but the virus must first spread to the older cohorts. And this lasts many weeks. And then they must come first to the hospital and then have a serious course of illness. And this lasts more weeks. And then there are other effects all of whose workings we could repeat. In the end, it won't lead to any new understanding of this phenomenon. We have no indication that the virus has changed. We expected that over the summer, there would be few cases. We know that in the summer, other coronaviruses don't appear, that the influenza doesn't appear in the summer. And therefore, we can't close our eyes to the fact that at some point, it will be winter again.


NOTES

1. A mnemonic for three pieces of advice to reduce the spread of the Coronavirus, namely, social distancing (Abstand halten), paying attention to hygiene (auf Hygiene achten) and wearing masks (Alltagsmasken tragen).

2. Actually two episodes ago, in episode 54.

3. In Niger State, which is in Nigeria, and not in the country of Niger.