I have friends asking why there is variation in the death rates by country for COVID-19, and if this is the worst pandemic ever seen? They ask, I think, because at my age, and with 45 years practicing medicine, I have seen a few epidemics. They are concerned that we may be overreacting if COVID-19 is not the worst. Others, usually former students, are puzzled by the epidemiology. Here is one perspective on their questions.
In 2009 a flu variant, H1N1, became pandemic. The death rates by country for this pandemic varied widely, with rates as low as about 2/10,000 in France (344deaths/2000000 cases), and has high as 19% in Spain (300/1578), with other countries varying from 6/1000 (China) to 6% (India). About 26 million tested positive worldwide. The US had 113,000 confirmed cases.
What about COVID-19? The death rate estimates vary similarly to the HIN1. For example the rates vary from a low of about 2/1000 in Canada (1/375) and Germany (14/6924) to a high of 7% in Italy (1809/24747). The country to country variations were actually greater during the 2009 H1N1 than with COVID-19. So, we can’t presently say that COVID-19 is worse than the pandemic of 2009. In fact, it looks epidemiologically similar. And I understand my friends questions; the 2009 pandemic did not evoke the same political and social upheaval in response. It is interesting to ask why.
There is another issue of comparison between the 2009 flu and the COVID-19 that informs the puzzlement about the reaction to our present circumstances. In 2009, H1N1 deaths were in younger people; 60-85% of deaths were in people under 65. Presently, COVID-19 reports show that the death rate rises as age rises. The average age of death in Italy, for example, was 81 years. Why excess deaths in younger people in 2009 did not spur a similar social upheaval reaction is puzzling. (Also, the usual seasonal flu tends to kill the older in excess, but H1N1 did not. The explanation is that the 2009 strain crowed out more usual strains. Maybe COVID-19 is crowding out H1N1)?
The variation in the death rates is likely due to an epidemiological error called, “ascertainment” bias, rather than anything to do with the virus. This bias arises when data for analysis are collected (surveyed, screened, or recorded) such that some members of the target population are less/more likely to be included in the final results than others. This bias messes up both the numerator and denominator for mortality estimates. For example flu tests were more attainable in 2009 than COVID-19 tests in 2020, changing the compared demography of who gets tested, how tests are conducted, and what populations are targeted for testing. This reason, bias, may be the likely reason for the varying death rates between countries.
I am pretty sure that COVID-19 did not dislike Italians more than flu liked them. Their reported death rate from flu was not as great as it is for COVID-19. The differences are likely epidemiological errors, in other words, bad science. But, bad science is the norm at the beginning of any epidemic. I just wish the public was being told of the biased uncertainty rather than being presented worst case scenarios as facts. “We really don’t know what to expect” is a truer statement than any present estimates being reported.
I wonder if the future title of the paper historically summarizing what happened with COVID-19 in 2020 will be titled, “The COVID-19 pandemic of 2020”. Or, “The epidemiological and public health error pandemic of 2020? Time will tell.
