Member-only story
The Death Rate of Covid-19 in Developing Countries
Why the pandemic has hit low-income places so hard
Throughout the pandemic, one of the hardest things to pin down has been the death rate of Covid-19. It seems simple — you just take your number of deaths and divide by cases — but in practice, we know that there are many people who are asymptomatic and will never get tested as well as other problems in counting the true number of people who have had Covid-19.
To remedy this, we’ve conducted detailed, careful studies looking into the death rate of the disease, using a measure called the infection-fatality rate (IFR). This is an estimate of the likelihood that someone will die of Covid-19 if they get infected, by age, and we are pretty certain of the results at this point.
However, for some time there has been an apparent contradiction. If you look at the death figures in areas across the world, some of the lowest-income places actually appear to be doing really well at first glance. A good example is India, where the death rate from Covid-19 is far below what you’d expect given the extremely high rates of infection.
So we set out to do a review of the Covid-19 IFR in developing countries throughout the world. Over the course of the last nine months, along with a team of researchers from across the globe, we’ve looked into the impact that Covid-19 has had on lower-income areas.
The results are, to put it mildly, quite worrying.
The science
Our new study — which we just put up as a preprint on medrxiv — builds on our earlier work that looked at the risk from Covid-19 in higher-income places, where we used estimates of the number of people with antibodies to the disease to work out the IFR, but includes so much more work. Because developing countries have issues with their death data, we spent a great deal of time hunting down the best estimates of Covid-19 fatalities in each place. Because antibodies decline, we looked at the specific tests used to determine whether people had antibodies to the disease and corrected for this in a…
