New studies seem to indicate that the COVID-19 fatality rate is grossly overestimated
When data on the new COVID-19 virus outbreak in Wuhan started coming out, the initial reported fatality rate was estimated to be around 3.4 % causing governments across the globe to panic. In the last couple of months, as the virus outbreak transformed into a global pandemic, and as governments rushed to contain its spread by deploying more and more tests, that fatality rate number kept being revised down. Towards the end of March an article on the Diamond Princess cruise-ship outbreak appeared in the highly regarded journal Nature, estimating the COVID-19 fatality rate could be as low as 0.5%.
In recent weeks, two new types of studies are shedding new light on the real extent of the people infected, and, subsequently changing our latest estimates of the fatality rate . The first type of these studies is using a more unbiased/random testing of the population, as opposed to only testing symptomatic people ( read in my previous article why this is important ). The second type uses antibody testing to test whether a person has already had the infection or not. All of these studies are pointing to current infected population numbers largely underestimating the real infection rate of the population ( the percentage of people infected by the COVID-19 virus ), which also means that we are grossly overestimating the real infection fatality rate. I have compiled here a few of these studies, many more are currently running in various geographical locations:
- New York, US: A study was conducted by The New York–Presbyterian Allen Hospital and Columbia University Irving Medical Center, where all women admitted for delivery were screened for COVID-1, 15% tested positive (33 out of 215 of which 29 were asymptomatic). When the study was conducted, the official reported infected cases in New York were around 1.3% — so underestimated by more than 10 x
- Sweden: A similar study conducted on pregnant women in Sweden, revealed an infection rate of around 7% , compared to the 0.1% of the population having tested positive ( 70x )
- Denmark: A study in Denmark conducted antibody tests on 1,487 danish blood donors, showing that the infection fatality rate of COVID-19 is closer to 0.16%
- Santa Clara, US: A similar study in Santa Clara was done with antibody tests suggesting that coronavirus infections vastly exceed official counts 50 fold.The study also suggests a revised fatality rate of 0.1–0.2 %
- L.A. county, US a very recent study shows that antibody testing in the county seems to indicate that roughly 4.1% of the population has already had the virus. An estimate that is 50 x higher than reported cases, bringing the reported fatality rate by that same number.
All of these studies seem to be converging towards some common facts. Firstly, the amount of currently infected individuals seems to be much higher than the official numbers show. With a large number being asymptomatic at time of the testing ( 85% in the New York study mentioned above ).This might also indicate that we are at a more advanced stage of the pandemic that we initially thought. An article just published by CNN reports that two Californians might have died of coronavirus weeks before the first case was officially reported. Secondly, the number of already infected individuals revealed by the antibody tests is also much higher, and seems to sit within the infection rates of the seasonal flu: 5–20% (US numbers, source). Last, and definitely not least , one thing prevails from these studies, and it is a piece of very good news: the real fatality rate is much lower than current official numbers show.
In light of all of these results coming out, governments and the WHO should investigate this evidence, and even run their own tests and experiments. If these results are validated, the WHO needs to publish revised estimates of the true infection and fatality rate of COVID-19, scientists should recalibrate their models and forecasts and governments need to review their policies in light of this new information.
Just to be very clear, I am not advocating that the virus itself should be thought of as another strand of the common flu, and treated as such. But I am saying that in light of fatality rates being revised down that dramatically and the total infected population being much higher than expected, that now might be a good time to pull the brakes on the general panic, even for just a short while, and to reassess the situation. We might also be able to answer some of the following questions:
- How far are we from herd immunity?
- Has the beginning of testing been grossly misinterpreted as the beginning of a pandemic that had started way earlier ?
- What does that mean for the predictive models used by policy makers ?
- What is the true impact of the policy measures that were put in place ?
I am following up this post with a more in depth article looking at some of the hard questions that we need to start asking ourselves, ranging from the effectiveness and long term repercussions of lockdowns, to the validity of the models used for forecasting and the ethical questions that we are faced with in deciding the right course of tackling this pandemic. Stay tuned.
Follow the author on @DonRiachi
