So I've been in a small kerfuffle with some folks on Facebook over the mortality rate of COVID-19. Their contention is that it is vanishingly small - like .002%. My contention is that it is much higher - like 5 - 6%. Since I've been told by some folks on FB to "do your own research," here it is.
---------------------------------------
Variables I use to compute the mortality rates:
Confirmed Cases: the reports of the total number of confirmed positive cases in the U.S.
Recovered: people who have a confirmed positive test and who have recovered from the virus (whether they were in the hospital, ICU, or not).
Deaths: well, yes. A confirmed positive case that leads to death. (Note that this can be tricky, since some death certificates which should read COVID-19 as the cause of death, instead read things like 'heart failure', pneumonia, etc. For example, my mother, who was a Type-1 diabetic, died of a heart attack. That's what's on her death certificate. But the heart attack was the proximate cause of her death. She really died from "complications due to Type-1 diabetes" which would have been a more accurate cause of death to put on the death certificate. The CDC and my other sources take these ambiguities into account when they produce death numbers.)
(Provisionally) Resolved: Since we're in the middle of a pandemic there are three types of COVID-19 patients. (1) those who've been confirmed positive, but whose cases have not yet resolved, (2) those who have recovered, and (3) those who have died from the disease. All three of these groups add up to the total number of confirmed cases.
Of those, the confirmed positive unresolved is the largest group; those are the people who have the virus but have not yet recovered or died. So we can't really count those people in a computation of the mortality rate because we don't know the outcome of their case yet. We can only count those whose cases have been "Resolved" in one way or the other. So we add those who've recovered to those who've died to get a single number of those whose COVID-19 cases have been resolved one way or the other.
This is why it takes the CDC several months to put out the annual report on seasonal influenza, how many cases, how many hospitalizations, and how many people died. To get a final number you have to wait until the number of new cases a week (or a day or a month) falls below a certain level, and all those confirmed cases have resolved before you can put out a final mortality rate for that year. For example, see the data for 2018-2019 here https://www.cdc.gov/flu/about/burden/2018-2019.html
So I use "Provisionally" Resolved because the number will change over time because we don't have the virus under control yet. This will cause the mortality rate to fluctuate.
Mortality Rate: MR = (Deaths/Resolved) it's the fraction of people whose COVID-19 cases have resolved and who have died from the disease.
=======================================
Here are the three sources of information I've been using in my research. I use them because they all have a history of doing this kind of analysis, they have been reliable in the past, and they are producing values that are pretty close together despite their being three separate organizations. Also, many other organizations (including the media) use their numbers when they report things to the public.
---------------------------------------
Centers for Disease Control - COVIDView
https://www.cdc.gov/coronavirus/2019-ncov/covid-data/covidview/index.html
and scroll down to find "Severe Disease -> Mortality"
The CDC doesn't seem to report "Recovered" cases, just Deaths and Cases. (But if you can find a CDC website that does include Recovered numbers, please let me know.) So there's no way to compute the "Resolved Cases" number and hence the Mortality Rate. So we just have to take their word for it that this is the correct mortality rate for this week.
For week 38 (14 Sep 2020 through 20 Sep 2020) "the percentage of deaths attributed to pneumonia, influenza, or COVID-19 (PIC) for week 38 is 6.6% ... the percentage remains above the epidemic threshold and will likely increase as more death certificates are processed."
The CDC also lumps in pneumonia, influenza, and COVID-19 as reported on death certificates into a single number. Given that it's September, the influenza numbers should be very low. I can't say anything else more specific about this data.
---------------------------------------
Johns Hopkins Coronavirus Resource Center
https://coronavirus.jhu.edu/map.html
As of 25 Sep 2020 about 2:23PM
Confirmed Cases: 7,019,232
Recovered: 2,710,183
Deaths: 203,329
Resolved (Recovered+Deaths): 2,913,512
Mortality Rate (Deaths/Resolved): 6.97%
This is the website that most people in the media and other independent organizations go to to get their data. These folks are really good at what they do and I'd trust their numbers any day.
---------------------------------------
Worldometers.info website
https://www.worldometers.info/coronavirus/country/us/
as of 25 Sep 2020 at 20:48 GMT (3:48pm CDT)
Confirmed Cases: 7,220,658
Recovered: 4,463,721
Deaths: 208,081
Resolved (Recovered+Deaths): 4,671,802
Mortality Rate (Deaths/Resolved): 4.45%
The folks at Worldometer have been around for quite a while and do a very good job of gathering and validating data. They also use a wide variety of sources to get their data. The folks at Johns Hopkins use Worldometer to help validate their own data. I'm not sure why their Recovered number is so much higher than Johns Hopkins, so that's an opportunity to do some more research.
---------------------------------------
So as you can see, it looks like the mortality rate for COVID-19 is somewhere between 4% and 6%. Compare that to the mortality rate of the seasonal influenza virus which is about 0.2%. So the novel coronavirus is between 20 and 30 times more deadly than the seasonal influenza virus.
4% to 6% is a far cry from 0.002%.
Be careful out there and wear your masks!
=======================================
UPDATE: (And thanks to my friend Mike Murphy for providing the links used here.)
First, an article at the Our World In Data website titled "What do we know about the risk of dying from COVID-19?" by Hannah Ritchie and Max Roser, dated March 25, 2020
Here's the link:
This is a good article that tries to define the different meanings of fatality rate from COVID-19 and clarify the definitions.
From that article I'll quote:
"When some people are currently sick and will die of the disease, but have not died yet, the CFR (Case Fatality Rate = the number of deaths/number of confirmed cases) will underestimate the true risk of death. With COVID-19, there are many who are currently sick and will die, but have not yet died. Or, they may die from the disease but be listed as having died from something else.
In ongoing outbreaks, people who are currently sick will eventually die from the disease. This means that they are currently counted as a case, but will eventually be counted as a death too. This means the CFR right now is an underestimate of what it will be when the disease has run its course.
With the COVID-19 outbreak, it can take between two to eight weeks for people to go from first symptoms to death, according to data from early cases (we discuss this here).
This is not a problem once an outbreak has finished. Afterwards, the total number of deaths will be known, and we can use it to calculate the CFR. But during an outbreak, we need to be careful with how to interpret the CFR because the outcome (recovery or death) of a large number of cases is still unknown."
This definition is kind of the closest to my Resolved Cases definition above and makes the valid point that the fatality rate will change as the pandemic progresses and you won't know the final fatality rate until the pandemic is over (which I believe I said above).
Mike also provides another link from Johns Hopkins on Mortality Analyses: https://coronavirus.jhu.edu/data/mortality?fbclid=IwAR2oFYxbghmty3mGM56I2HYRc6QQtCqu9V6CbXiu5BPf6zYQLWFNJtVZOtg
The Johns Hopkins page gives several good graphs on, in particular, "case fatality ratios (the number of deaths divided by the number of confirmed cases)."
Note that this is different from my fatality rate which is "the number of deaths divided by the number of RESOLVED cases." As noted above, I use RESOLVED cases because the pandemic is still ongoing and many of the current confirmed cases have not yet either recovered or died, so it doesn't make sense to me to include them in the denominator.
Finally, Mike shares a link from the research journal Nature, dated 28 August 2020, https://www.nature.com/articles/d41586-020-02483-2?fbclid=IwAR1r5WxYkOj8lm8qkOB7uvpgFiR3_X0IfArw2eCMR8vwBJE2Y1OR0BDtYqw
The article is titled "The coronavirus is most deadly if you are older and male — new data reveal the risks" and is by Smriti Mallapaty, who is the Asia-Pacific reporter for Nature News.
Mallapaty talks about a ratio "a metric known as the infection fatality ratio (IFR), which is the proportion of people infected with the virus, including those who didn’t get tested or show symptoms, who will die as a result." which is close to one described in the Our World in Data article described above. The main point of the article is that if you are male and/or over 50 your chances of dying from COVID-19 are way higher than if you're under 18. Something I think we've known for a while.
Aside from the definition of infection mortality ratio, this article really doesn't have much to do with the discussion in the original blog entry I wrote above, so we'll just leave that here.
---------------------------------------
