At the beginning of the coronavirus outbreak, fatality rates were based on those testing positive for the virus using mucus-based swab testing. Now, we have the results of an increasing number of community-wide antibody blood tests, and they paint a different picture.
The main SARS-CoV-2 test involves taking a mucus sample and analysing it to find traces of viral RNA specific to the virus, a process known as reverse transcription polymerase chain reaction, or RT-PCR. These tests, which test for the presence of the virus, have typically been rationed and used mainly on those with obvious symptoms.
Early on in February 2020, based on data from Wuhan, the World Health Organisation (WHO) reported a crude fatality rate of 3.8%. Later on 3 March 2020, it reported that 3.4% of those infected had died.
As the world’s understanding of the disease progressed, it became clear there were significant numbers of infected people with mild or no symptoms and many began to question early fatality rate estimates.
Antibody tests, rather than testing for the presence of the virus, test for echos of its past presence in the form of specific antibodies. They test who has had it rather than who has it, and are particularly useful for detecting past infection in asymptomatic cases.
Only when we have a more accurate estimate of how many have been infected can we meaningfully estimate the infection fatality rate (IFR).
This week the results of the largest antibody study so far were published in Spain. The Spanish study, which took blood samples from nearly 70,000 people and tested them for SARS-CoV-2 antibodies, suggests around 5% of Spain has been infected, a figure 10 times higher than the 0.5% based on the official number of people testing positive on RT-PCR tests.
5% of Spain’s population is around 2,350,000 people. With deaths of 27,400 this study suggests the IFR is 1.2%, a figure far lower than the WHO’s early estimates.
Testing in Geneva suggests a similar IFR. An antibody study in Geneva suggests that 9.7% of the population has been infected. This figure yields an IFR of 0.5%. However, looking more closely at the data, it appears Geneva had an oversized share of infections among those under 50, an age group that has suffered 2 deaths.
Among those 50 and over, 267 have died and 3.7% have tested positive for antibodies. If we somewhat crudely assume that no one under 50 dies from the disease, because so far only 2 out of 269 have, and apply the 3.7% to the total population, we get an IFR of 1.4% (1.4% = 267 / (3.7% of 506,000)), a figure broadly in line with IFRs based on the studies in Spain and NYC.
Taken together, these rates of 1.2% (Spain), 1.2% (NYC) and 1.4% (Geneva adjusted) suggest the real fatality rate is far lower the WHO’s earlier estimate of 3.4%.
The recent Spanish study also offers clues for discerning whether or not you have had Covid-19. 43% of those losing their smell tested positive for IgG antibodies. This was the most telling symptom. Those reporting no symptoms made up close to a third of all of those testing positive for antibodies, a finding in line with other studies that find a large number of asymptomatic cases.
Curiously, 17% of those with a positive RT-PCR test did not test positive for IgG antibodies. It is not clear why. It could be a testing issue. It might also have something to do with the antibody response and its timing.
Research paper (in Spanish)