 For more videos on people's struggles, please subscribe to our YouTube channel. Welcome everyone to today's discussion. We are going to discuss today COVID-19 related death estimates in India. It has been under a lot of controversy. And what we have with us is Professor Christof Gulmoto. Professor Gulmoto is a senior fellow in Demography at the French Institute, IRD. He has joined the Center for Science and Humanities in Delhi in the year 2021 to work on demography of inequalities. Recently he has contributed a paper looking at critically reviewing India's death estimates and providing alternative estimates. Welcome Professor Gulmoto. It's a great pleasure to have you here. So as we discussed in and this is very much in the news currently in India that India's COVID-related death estimates are being debated a lot in the country. Bodies were floating on the ganges. We saw scenes of crematoriums overflowing with people and that shook our conscience, the nation's conscience in a great way. There have been reports that the government estimates of deaths are much underreported compared to what could have happened to the population. It's very important that we record deaths and we have proper death registration. Professor Gulmoto's own estimate suggests that we had 2.2 million deaths in India till May whereas official figures are only 0.4 million. Professor, your initial reactions, why should we have a better estimation of deaths, why death registration becomes very critical to a health system or country, not only in the context of pandemic but in general, your initial thoughts on the current situation? Well, I think the first issue is to count people and to give recognition to people's lives and that is mostly about their births and their deaths. And as you probably know, there's been a tradition in Western countries related to the Church of keeping record of deaths and births and marriages as well for more than 500 years. Whereas in India, it's pretty new. It dates back to the colonial period. Of course, death is taking but also birth and death registration which was introduced in the 1870s in some cities like Mumbai and Kolkata. But strangely, this process of registration of death and birth has moved on very slowly including during the last 50 years because there is still serious under registration of birth and death and especially of death. So birth as we know are more often registered because a certificate of birth is useful for many purposes especially for schooling the children. But death is considered of lesser importance to register even though it is supposed to be useful to sort out inner-reference issues and many others. But first of all, death is locally recognized in communities and that's enough to decide about sharing of property and or residential change. But otherwise, the under registration of death is still very important. It's still very significant in India because a lot of people do not consider it necessary to go and report death. So it varies in some areas and the registration of death is minimal. It can be for example in some cities or in Kerala or in Punjab. But elsewhere like in North Central India, about half of deaths are actually reported. So that's the first environmental feature to take into account. And then there are sudden deaths due to a catastrophe and that is to some extent something new in India because there's never been a catastrophe of death volume in India for the last few years maybe since the Bengal famine. So now there was a need to quantify this catastrophe and to quantify it just to let the country, to less citizens understand what has happened. And I think one thing that is very important and it has probably impacted a lot of us resident in India is to see and to hear about people dying around us. So that's new to me. You know, I'm coming from France. We had two serious waves. But to tell you the truth, I had nobody close friends or relatives who died. Whereas in India, after two months, it turned out that had several friends who died. So I realized that finally this crisis was for real beyond numbers. And then I asked for some of the friends here in India and I don't have so many. So that's why I was surprised that some of them, some of my own friends died. Sorry, but also surprised that it could happen. Then I asked other people and everybody knew of someone. So sometimes it was somebody close enough among their relatives. Or anyway, it was somebody they knew, at least one person. And I said, then how is it possible, you know, from a statistical point of view, how is it possible that everybody knows at least one or sometimes more people affected. And here I'm not talking about somebody who got COVID, but somebody who really passed away. And that's, I think the beginning for everyone in this country is that it's not only about what newspapers are saying, it's not only what numbers are being published. It's something that is affecting us, you know, be it in South India or in North India everywhere. This seems to have happened. And I think for most of us, it's something entirely new. We've never heard of something we've really heard during the last 50 years of people who might have died of, say, road accidents, but not within a year. So this became something considerable in our lives. And so it should be, at some point, be translated into numbers. Definitely. I think you made a very profound point. And that's very core to our understanding. Everyone in our family, every one of us have someone in the families, relatives, friends dying in this period of time. And definitely the government estimates currently are not able to reflect the magnitude of the problem. So coming back to our issue of estimating COVID-related mortality, there are some systemic issues of underreporting because of weak system, weak registration system. But also there are some, I think that there are systematic ways of undercounting COVID-related deaths. But how can we, how can we have better estimates, better methods of estimating COVID-related mortality if we do not have good registration system? Or, say, our health system is not equipped to capture what could be the other ways if you can enlighten us on the methodology of it? First, I guess we need to understand what's happening after a death. So it may be a COVID death, COVID-related death. It may not be. So if it is not a COVID-related death, let's forget it. It can be, for example, road accidents. So there is no doubt. But then there is a huge category of deaths which were of unknown reasons, especially, I would say, in rural India. And again, even local people had no clear idea whether it was a sudden fever or something like that. So there are no words to describe it. They might have heard about COVID, of course, but it didn't relate to their past experience of morbidity, of disease, of bad health, and finally of, and so a lot of these unknown deaths were just ignored. And some of them were due to COVID. And at times, even people had the fake feeling that this death may be concealed because they were, the family could be looked down upon as a family with the disease. And so they better attribute the disease to something else like all the age, because mostly all people were affected by COVID. So all the age is a blanket answer for death. And then you had deaths that were, in fact, clearly identified or could have been identified as COVID deaths. Now, a lot of these deaths are simply not registered. And so even if they were properly identified by the relatives or even doctors, but anyway, they were not registered. And so among those who were registered, there was even another difficulty is that they needed formally to be tested for COVID. But a lot of people, I mean, we've heard stories of people dying before getting even tested, and that in cities. So you can imagine that in rural areas, people were just, there was no way for them to test them in a primary health center or still they were dying. But then they would not count as COVID deaths because, and then finally, among those who were tested for COVID, some of them, and we don't know what proportion, in fact, some of these deaths were attributed to co-morbidity to some other thing that people would die of, say, diabetes. We know that diabetes patients were extremely vulnerable to COVID. So the cause of death would be attributed to something else. And this is definitely not what it should have been, what should have been done, because right from the beginning, the WHO requested practitioners and registers to use a very broad definition of COVID deaths. So whenever there was a suspicion of the deaths being caused by COVID, it should be registered at COVID, by COVID, and not under some other category. So here we are talking finally about the minor parts of deaths which were identified as COVID. A lot of them were just ignored. We, nobody had an exact clue or wanted to come forward and say it must be COVID. Among those few that identified, then they had to be tested and most of those recorded had to corresponded to persons who died in a hospital. And we're finally there. Somebody could say, yes, it is COVID and it is not another source. So if you look at the long line of the different steps whereby deaths may be ultimately registered as COVID deaths, they are, I mean, every time you go from one step to another one, being identified, being registered, being tested, blah, blah, blah, you lose 20% or 30%. We have no clear idea. But what we discovered soon that beyond our experience, which told us that it must be more than the official numbers, beyond that experience, we've seen local numbers that did not match with what people were observing on the field and be it registered deaths, be it deaths of all kinds. And then later on, incinerations or other dead bodies identified the number of which would not match official numbers, which I wouldn't call by the way government numbers. I would say just official numbers because they were counted by different local health authorities. And more or less everywhere we had the feeling that there was severe underestimation or say under registration.