Confusion Reigns about COVID 19 Fatality Rates
This is my final post on this subject—at least for now. I am disappointed by all the confusion evidenced even among very bright people—including some who have responded here.
The confusion is over two very different statistics that people cannot seem to keep distinguished from each other. One is the true IFR (infection fatality rate) of COVID 19 and the other is the CFR (case fatality rate) of COVID 19. Now, to add to the confusion, some reporters of statistics are calling the IFR the “IFR-S” which is the same as CFR! (It means infection fatality rate among those with symptoms of the disease.)
Recently someone here challenged the Heinsberg Study that I reported on and said the U.S. news media should be reporting about. The key value of the Heinsberg Study (out of Germany) is that it sought to establish, as much as possible the true IFR, not the CFR (or IFR-S). That person (I don’t remember who) referred me to a study conducted by the Johns Hopkins (university) Repository that was reported recently on the front page of the New York Times.
Please (!) people. Read carefully that report was about the probable IFR-S (which is the same as CFR!). The researchers who came up with the probable IFR-S (CFR) admitted that the 1.3% number did NOT include asymptomatic people. It obviously also did not include people who may have had symptoms but never went to a doctor or were tested but just stayed home and “rode it out.” We are discovering that number is higher than anyone suspected months ago.
So, my point is that the Johns Hopkins study reported on by The NY Times is NOT comparable to the Heinsberg Study and others that focus solely on the infection fatality rate and project what that probably is based on research on a population of diverse people. (Yes, the “carnival” included elderly people and all kinds of people representative of Germany society generally. I have been to such “Fasching” carnivals in Germany and know that everyone who can attends it.)
If you want to read more about the New York Times-reported Johns Hopkins Repository study go to Study Puts U.S. COVID0-19 Infection Fatality Rate at 1.3%. You will see that the reporter (John Commins) admits that what the study REALLY showed is that the IFR-S (Case Fatality Rate) is probably around 1.3%. Still much, much lower than thought earlier. And keep in mind that the IFR-S is NOT the same as the IFR.
Now, look at this article by epidemiologist Gideon M-K (April 27): What Is The Infection-Fatality Rate of COVID-19? (Google it.) This researcher studied numerous studies from around the world and concluded that the real IFR is probably somewhere around 0.75%. That is, probably, about 7.5 out of a thousand actually infected people will die and that takes into account vulnerable people.
My point here is that everyone who has any interest in these statistics MUST learn to recognize the difference between CFR (which by all accounts is quite high) and IFR (which by most accounts now is much lower than the CFR). And recognize when IFR-S is being used which is NOT the same as simple IFR but closer to, if not the same as, CFR.
Now, I’m not as interested in the statistics (which no doubt will change) as I am in the difference between CFR and IFR—which is a crucial difference. Obviously CFR is going to be higher now that we know many people who acquire the COVID-19 virus do not get tested and are never diagnosed.
Final note: Of COURSE the IFR is going to be higher within a nursing home where the virus has invaded. The IFR being sought and estimated is the OVERALL IFR for a large population such as the U.S. or a city, county or state (or other country and sub populations within it). The reason this is important is that it makes a great difference when “re-opening” economies and how to do it. A very high CFR or IFR-S is not really helpful. What we need to know is the likely IFR.
Note: If you choose to respond, be sure to stick to the subject, do not go off on tangents, do not use my blog to write your own blog post, do not rant, be civil and kind, do NOT include any hyperlinks and be brief.