 Hello and welcome to NewsClick. Today we have with us Prof. Satyajit Rat and we will be discussing COVID-19 and the various facets that is beginning to unfold. Satyajit, we have had this rather interesting article from the economist which has looked at what is the growth of numbers which they think have taken place as against the registered cases and therefore, how far are we away from, for instance, herd immunity, what are the likely trajectory of the disease and so on. But interestingly, the graphic also tells a story about India that India had very large numbers according to it quite early. March, April, they think the numbers were quite large based on the seroprevalence studies which we have already discussed. Now you see doctors seroprevalence studies, now the studies were adopted, the results were sort of selectively published and hotspots not published. Looking at it, what is the first, what are the impressions about what the economist has done, how correct is that and then we can talk about the prognosis that they have also talked about. Okay, so first off, right away, the economist is both right and wrong. You realize that this is what many of us say quite commonly about the economist. But the economist is both right and wrong. It's right in that the broad contours of the calculations that they've done are reasonable, sensible and we can talk about them in a minute. But I want to start out by underlining what I think is wrong. And what is wrong is that the economist article is actually averaging across massive variation. And averaging across massive variation gives you wildly erroneous estimates potentially. So you're talking about the ICMR is selectively cherry picking seroprevalence data from their April-May surveys. And one of the problems that we discussed at that time was that, in fact, they were taking samples in areas where the seroprevalence was low, they were excluding so-called hotspots and if they had taken those together, then the averaging would have looked completely different. But the trouble is that that average would not have been representative because we didn't know how many other hotspots there were. And it's in this sense that the sheer variety of an emerging pandemic make, let's make no mistake about it. This is still an emerging pandemic. We're not that far into the pandemic for us to say, oh, now everything is settled. It's not settled. And therefore, it's breaking out here, there, everywhere. And therefore, the sheer diversity within a community is so large that taking the small scale seroprevalence studies that we have and using them applied across boundaries of political geography is laying ourselves open to massively erroneous estimates. That is the real limitation of the economist's analysis. I wanted to interrupt you here for a moment that actually there are two levels of misconception here. One is that even if you do seroprevalence studies in certain districts, for instance, and that's what the Indian study was, but similar studies have been done elsewhere too, that there are two levels at which you can make mistakes at the level of the communities, even in an area that the seroprevalence study would show very different results if it is done uniformly across the area that it is done in sample basis, which is what was done. And secondly, the areas themselves being chosen are also not representative. And that's the nature of what you call the emerging pandemic that it moves from area to area from pocket to pocket. And there is no, in fact, statistical way of saying if I have calculated x number of samples, it will be representative of the country unless you really take a very large sample from every possible scaled size of the area. That means you take it at the village level, in the village you take it within the each locality level and so on. And that's a scale at which it really cannot be done. And even then you have to take it every month to see how it's moving. Exactly. So that's really the limitation of the economist analysis. But that said, as I said, the broad trajectory of their analysis is unexceptionable. It's quite reasonable. And they are saying things that, in fact, most analysts have been saying for the past few months, as a matter of fact, we on our weekly discussions have made many of those points ourselves. So none of the economist's analysis is a surprise. But if I may, let me use the hook of the economist's analysis on a global scale and turn that around and use three completely isolated examples of very small, relatively very small scale surveys that converge very interestingly. So those three zero surveys have been reported, two in preprints and one in an article in the New England Journal of Medicine over the past couple of weeks. One is in the ski resort town of Ishgl in Austria. One is in the entirety of Iceland, which is really not even Daravi size. And one is in the Brazilian Amazon, as the authors call it, although in reality in the cities of Manaus and Sao Paulo. And for these three widely disparate locations, we have remarkably interesting data sets. We have the zero survey. We have the number of so-called hospitalized cases, which is the less interesting number. But the even more interesting number is we have fatality numbers. And this is a situation in which we can actually extrapolate from the local zero survey to calculate the number of people infected. And because there is an estimate of the number of fatalities, we can actually calculate the infection fatality rate. Okay. So what you are also implying, I guess, is that the fatality is a hard statistic while infection, disease, all of them are relatively more difficult to really get as a hard statistic. So I'm going to be irritating about this and say yes and no. Okay. So your point is your question is absolutely yes for the Austrian location and the Icelandic location and not quite for the Brazilian locations. But here's what happens. The remarkable convergence and it's quite noteworthy that the infection fatality rate calculated for Ischgl in Austria is about 0.3 percent. The infection fatality rate for Iceland is about 0.3 percent. The infection fatality rate, if you take virus-detected fatalities in Manaus and Sao Paulo is 0.1718 percent. But the authors point out that a large number of deaths that otherwise looked exactly as though they were COVID-19 deaths, but were not tested for virus are included. Then the infection fatality rates both in Manaus and in Sao Paulo come to about 0.25 to 0.28 percent. Okay. This is a remarkable convergence of infection fatality rates across. I'm going to again interrupt for a moment and say that there is two things we have to also tell our viewers. There is an infection to fatality rate that we are talking about, but the case fatality rate which we have earlier talked about is those who have the symptoms of the disease and have been identified with the disease and case fatality and infection fatality are two different things. Unless you're going to tell me yes and no again. Well, unfortunately I am because health authorities across the world over these past few months have been constantly redefining what they mean by a case. Okay. What you have said is what we should be doing. The WHO still say this is the way to look at case fatality ratio. Yes, but the difficulty is WHO is actually not defining what a case is. That is also true. And the trouble is that health authorities across the world started out by using the definition you pointed out, which is people who feel that they are unwell are cases. The trouble is everybody who was virus positive when they were found in contact tracing are most of them, if not all the way, most of them have no symptoms, but they are being recorded by health authorities as cases found. Okay. So the idea of the case fatality rate for me has steadily lost ground and relevance because it's not comparable from time to time to time. From March to August. Precisely. This is why the infection fatality rate has reflected by some reasonable cumulative evidence of infection prevalence by antibodies and CRO surveys and the cumulative evidence of recorded deaths with all the errors inherent in that is still the number that I would prefer to be looking at more and more and that number in these three studies is quite remarkably overlap. But I wanted to again do a sideways leap over here, but when we talk about for instance the influenza cases, we actually talk about not the antibody, antigen response, but we actually talk about the people who have the disease. And then when you talk about the case fatality ratio in the case of flu, which is less than 0.1%, it is really those who have been identified with the disease. Well, epidemiologists might disagree because again, even for influenza, different health authorities do different things a little bit. But let me reinforce your point by pointing out that even if you simply take the 0.05 to 1% number, for whatever it is worth for seasonal influenza, versus a 0.25 to 0.3% number, infection fatality rate for COVID-19, that's three times as much number. But let me add to that, there is steadily growing evidence that unlike influenza, a very substantial proportion of COVID-19 infected individuals do not recover, they don't die, but have protracted illness. How protracted? We don't know. But I am beginning to see numbers from around the world and at least 20% of people are still sick more than two months after detection of infection. That's very unlike influenza. And secondly, influenza, those who are infected but don't show disease, do not infect others in the same way that COVID-19 seems to do. So the transmissibility of COVID-19 seems to be also therefore, relatively much higher and it's not seasonal unlike flu. So there are other issues too. So you're absolutely correct. In addition to the infection fatality rate difference, the point that you make about the noisy patchy inefficient transmissibility of influenza infection as opposed to the rapid spread of COVID-19 is a major issue for communities, as is the third point we just discussed, which is the potential long-term consequences in a fairly substantial proportion of infected individuals for COVID-19. All of that together makes COVID-19 nothing like simply again to use the inimitable idiocies of the leader of the free world, something more than just a strenuous flu. Which we now know he really knew wasn't true. So it was not ignorance, but it was basically propaganda if you will, if you put a best spin on it. Coming back to the issue that you're raising, you can go back to the economist article. So what you're saying is broadly the numbers in spite of all the problems that they have are ballpark-wise not very different from what perhaps really happened. Of course, we will know that if we do much more extensive sero-service, you might have better figures, but this seems the ballpark figure seems to be right. Is that what you would say? Very reluctantly. Give or take a factor of 10 maybe. Give me a fairly large uncertainty. Okay. So they have said 600 million, odd million could be infected. I guess 30 million who we know or we have identified. So it could be 20 times. It could be 10 times. It could be 30 times. We don't really know. But that's really the kind of uncertainty we are talking about. But it still means that a significant part of the people are still without protection. Even if we accept that if we're infected, probably the immunity will stay with you for a year or so. At least for a year, maybe two years, we don't know. But probably about a year or so at least. Yes. And again, the Iceland survey that I mentioned actually suggests that unlike some other sero-service, they don't show as rapid a loss of antibody levels. So I suspect again that between communities, between different parts of the world, there are going to be differences about how long natural SARS-CoV-2 infection immune responses at least at the level of antibodies last. And it's possible that the data are going to be equally noisy in that context as much as they've been noisy in other contexts. But yes, I would guess between six months to a year is quite likely. So some total of it is still irrespective of all these results. We still need the vaccine if we want herd immunity. Given that it's not therefore a small flu, we have already crossed a million deaths. So it is something much higher than any other disease, except perhaps the annual death of tuberculosis, which also includes a lot of HIV patients who died because of TB. So if you take all of that put together, this is continuing to be in seven months, one of the highest mortalities that we are seeing due to infections. So this is still going to be a major problem for the world unless we have the vaccines. Absolutely. And in that sense, let me have an unqualified good thing to say about the Economist article. Their analysis of excess deaths is quite spot on. And their analysis suggests, in fact, that even in prosperous countries, there may be a substantial undercounting of COVID-19 related deaths. They suggest, I think if I am remembering correctly, a global undercounting of by about a third, meaning that they think that there are 50% more deaths than we've counted so far. That's an undercounting by about one third. I think that therefore, the undercounting of COVID-19 related deaths is quite significant and is going to be even more significant in poor communities than in prosperous communities. And I think therefore that that component of the Economist's analysis is very valuable in making the point that you're making that this is a truly substantial significant problem. Remember the dramatic point that the Economist makes. With respect to the United States of America, the Economist points out that COVID-19 deaths in the US in this calendar year have been half of the number of cancer related deaths in the calendar year and that will be expected from cancer related deaths. And the year is not done anywhere near yet, as the Economist points out. So we're talking about a very large disease and death footprint across the world for the pandemic. So absolutely, we need vaccines. So coming back to Indian data, we know the number of places it was, if it was, you had other co-morbidities and other problems, quite often COVID-19 was not stated in the death certificate. Of course, even the figures that the government has given, it does not actually tally in number of places with the figures we are getting from the crematorium or the burial grounds. So the all kinds of mismatches. And we have to also really think about this, that because of COVID-19, people are afraid to go to hospitals, go to hospitals later than they otherwise would. And some of the hospitals without, for instance, COVID-19 certificate will not admit you. So they're not geared up to take patients who have COVID-19 and therefore they say prove that you are not COVID-19 affected when you are really in an emergency situation. So we have all kinds of other issues also, by the number of deaths, which are not directly COVID related, but are related to the public health crisis that COVID has caused. So that is also something that we have to take into account in particular in countries like India. Coming back to this, and we don't really have the time to discuss in detail the vaccine issue, but Punawala has raised a very important point. He says that the vaccine costs are going to 80,000 crores is government of India prepared for it. He's not given a breakup of the 80,000 crore cost. But if we do a simple back of the calculation that even at $3 in a vaccine shot and it's a two shot vaccine in his case, oxford vaccine, we come to something like 60,000 crores. Now, if you take the cold chain and giving the shots, all that delivery costs into account, you're really not talking about something like 150,000 crore a year in cost for providing the vaccine to all the Indian people. Now, is there any preparation that we see or we know that the government is doing on this count? Well, we've discussed this earlier that there is a committee that the Government of India seems to have formed under the chairmanship of a Nithya Aayog member for planning implementation of a vaccination program. But what this committee has thought, I have not seen reported. So, unfortunately, no, we don't know what the government is planning by way of implementation. This is particularly worrisome because a lot of the vaccines are going to require stringent cold chains. A lot of them are going to require second doses. Many of those second doses are differently spaced. So, for some of these vaccine candidates, there is a three-week interval. For some, there is a four-week interval. Just the fact that there may be multiple vaccines complicates a vaccination implementation program immeasurably because how do you decide whom to give what? Which vaccine? Are you going to implement it region by region? This is a logistical nightmare that needs to have been thought about by governments much more granularly and much more transparently than we have evidence for so far. This is, of course, about the Government of India. But I actually don't think that any other government in the world has done particularly better at this. In fact, I'll end with this, the Government of the United States, where it says it's going to start the vaccination, and this is what needs to be done. The state authorities there, this is also a federal country, said, hey, you haven't told us anything about this, and now suddenly you say we have to make vaccination programs, cold chains, and some of the cold chains are not like what we are probably going to have as vaccines here, but some of their cold chains are minus 72, minus 80 degrees centigrade cold chains. So if they are scared of how they are going to do it for a smaller population with much more resources, with a much better refrigeration, freezer system than we have, this is really going to be a major issue. And just for the viewers, we have all Spolio programs have at best done 130 billion or 13 crores per year. And we are really talking about something like 100 crore per year, if not at least 70 crore per year, and don't forget that even by the best criteria that we are now seeing is a 50% protection. So really are talking of something of major dimension, and not one dimension, but multiple dimensions for which we should be transparent with how we are going to do it. Otherwise, I don't see it being successful, particularly as this government has centralized itself so much that participation in the states and local authorities is going to be absolutely necessary. And the public health community, none of these are in the offing as of now. Absolutely. So and all of this is in the background when I say that the numbers of how many thousand crores it will take for a vaccination program that you referred to as Mr. Aditya Poonawala saying, I think those are just shots in the dark. I don't think we have any realistic estimate of what it will cost because we have no estimate of how we are going to do this. Thanks Aditya for being with us. And we hope that at least the good sense will dawn on the government that they don't start vaccination using the local farmers and the police stations like they have implemented all public health measures till date. This is all the time we have in Newsweek today. Do keep watching Newsweek and do visit our website.