I’m a data analyst. You can’t use recovered cases because it is meaningless in this case because you can catch it again. Recovered just means you have re-entered the subject population. You use confirmed vs deaths because deaths are a subset of confirmed. Deaths and recovered are mutually exclusive. You cannot be dead and recovered so it doesn’t make sense to use them in a ratio.
I will disagree with the assumption that you can catch it again. Scant evidence as to that
Yes you’re a data analyst but not a medical person that has to work with numbers the way I have to, speaking with competent doctors who explain this stuff to me. They are scared!
There is no proof that people are “catching it again”, they may be relapsing because they never cleared the virus in the first place! Even regular influenza folk thought to be on the mend relapse. Even so the number of such covid relapses so far in terms of recovered persons is very low and whether or not such cases are considered as a separate case may be subject to each nation’s reporting requirements and are part of the statistical noise we find beyond the decimal points in the 100ths and thousandth’s places that we round off from to get our measured percentage points rounded to a a 10th’s place.
Even if they ended up being reinfections, they are low in number and they might still be considered as a new case with new dynamics which will either progress again to recovery or death.
In other words reinfections wouldn’t change the way the math should be figured depending on how they were counted unless the numbers become significant in the thousands...but there again...had they really recovered?** If not, it would be logically consistent to subtract them from the numbers of resolved cases and list them back as active ongoing cases again.
The bulk of recovering persons will stay recovered, the death rates will pad downward. Again I say the US data reflects an abrupt upward shift in existing corona cases but only because it reflects an increase of testing of cases already suspected and finds ongoing cases that have been goin on for the past 5 to 20 days.
How that shakes out in terms of actual need for hospitalizations, we won’t know for at least a week or two. If hospitals actually start getting super slammed then we know it isn’t just an anomaly of early useless data but that the virus is spreading. I expect an upward spike in positives, when the superslam occurs, beyond our early test data. If there is a capping off of ongoing cases and a rough equilibrium occurs of case closures to new diagnoses(dead and recovered) then we’ll know that the disease has hit a peak and we should soon start to see more case closures vs diagnoses.
Hopefully, those who have recovered will depress the percentage curves of those who will have died. When the statisticians 100years from now measure the TOTAL CASE rate from the final death rate of the great 2020 corona virus pandemic, the final mortality rates might just be be in that 1.3 percent rate you have touted (though hopefully much less)...unless the American total case rate tops 10 million persons....then in that case 130000 souls will have died of it.
**Criteria for reinfection for covid don’t quite exist unless we are testing for sub strains that are different than what infected one before. The suspicion is that a person seemingly getting it again within a week or two prior clearance may not have actually cleared the virus. It may raise the possibility that the virus may become latent in some folks the way Hep c, hep B or Epstein Barr virus does; if that is found to be the case as they study this virus(remember all the research is new and ongoing....we don’t much about long term after effects yet)then the concern will be if such persons will potentially be aerosol spreaders of the virus for the rest of their lives...
kind regards....
Ps: Don’t count the China numbers...https://twitter.com/jenniferatntd/status/1241082360323596289
A lot of Chinese phone numbers seem to have gone off line in the past 2 months...15 million in fact!