March 12th 2020
This morning, we learned that actor Tom Hanks and his wife have contracted COVID-19 infection. Indeed, 43 famous persons have already been affected by the disease, including six Iranian leaders and four European soccer players. We might speculate that this reflects a single exposure event in Iran…or the fact that European athletes travel frequently in a high-incidence environment. Perhaps similar reasoning can be used to explain the striking variation in coronavirus death rates between countries.
As of March 12, 126,258 cases of COVID-19 had been reported worldwide; and 4,368 died of the disease – a case-fatality ratio (CFR) of 3.6%. This figure varies widely from country-to-country. Indeed, The CFR in Mainland China is 3.9% – vs. 3.2% of all other countries combined. Only 1.0% of infected passengers aboard the Diamond Princess cruise ship died of the disease. Among countries reporting more than 500 cases to date, the CFR (in descending order) is Italy 6.6%, Iran 4.3%, Spain 2.8%, United States 2.8%, Japan 2.5%, France 2.1%, United Kingdom 1.7%, South Korea 0.8%, Netherlands 0.8% and Switzerland 0.7%. Fatal infection is notably rare among the Scandinavian countries: Sweden 0.2%, Norway 0%, Denmark 0%. Finland, which has experienced 109 cases, has reported zero mortality.
The CFR in each country is largely determined by how the disease is identified, defined and reported. The quality, training professionalism and available resources available Health-Care workers will also influence case-finding and treatment; as will Demographic and cultural differences related to age, nutrition, access to local health facilities, lifestyle and exposure to animals. I suspect that much of this variation in CFR for COVID-19 is related to the very definition of “cases”. The death of an infected patient will be obvious and easily documented; while asymptomatic or relatively minor infections could remain undocumented. Indeed, the total number of “cases” used to calculate CFR might only represent those cases who are sufficiently ill to seek medical care. The patient with a mild febrile illness will not be “counted”
These questions can be easily solved using a standard serological survey in relevant communities. Such a survey should include a questionnaire regarding recent symptoms, exposure, occupation, etc. If a large proportion of the general population is found to be seropositive toward SARS-2 virus, we might conclude that the disease is less dangerous than current statistics seem to indicate.