Replies

zeestephen,

The case fatality rate is going down in the US because the number of cases is going up. This is clearly partially due to expanded testing, but it also has largely been to expanded transmission reaching the symptomatic phase - the fundamental problem with using this statistic to project without cohort information.

Here is an case-fatality example of a country with comprehensive testing past the rapid-growth part of the bell-curve:

On March 6th the S. Korea death/cases rate was 0.6%. It’s continuing to creep up as people have time to die. 57% are still sick.
3/22 8897 104 2909 1.12%
3/23 8961 111 3166 1.24%
3/24 9037 120 3507 1.33%
3/25 9137 126 3730 1.38%
Projecting from this end, the current trajectory is a final number around 2.0-2.3% if no miracle cure, and no significant new cases.

//
FWIW, here’s China’s numbers:

Do you recall when the case rate for China was solidly below 2%?

I did not record the initial figures, so I only have them at hand since 3/5.

Date Cases dead recov dead/case
3/05 80422 3013 52229 3.746%
3/08 80735 3119 58588 3.863%
3/10 80956 3162 61559 3.905%
3/13 80945 3180 64194 3.928%
3/16 81032 3217 67910 3.970%
3/19 81155 3249 70535 4.003%
3/22 81409 3274 72808 4.021%
3/25 81726 3291 74173 4.027%

This has a trajectory to about 4.2-4.3% as a final rate if nothing changes. That will need to be offset by earlier suspected unknown cases, which looks to take it perhaps to about 3.3-3.7% ...but it is China, so caveat emptor.

3/9, Germany was being touted because it had no deaths at that relatively late point. In this case the rise is on a different end, because while it had these cases, there had not been enough time to die. BTW, the demographics here are entirely different from Italy, as the infections at least are heavily young athletic people who ski.
3/9 Germany has 1151 cases, no deaths, and only 18 recoveries.
3/22 24873 94 266 0.377%
3/23 29056 123 453 0.423%
3/24 32781 157 3243 0.478%
3/25 37323 206 3547 0.551%

“Just be scared all by yourself. How is it that Germany has 800 cases and zero deaths? Other countries are similar. Norway, Sweden, hundreds of cases, no deaths.
It seems certain areas and people are much more susceptible than others.”

51 posted on 3/9/2020, 10:42:18 AM by SaxxonWoods

Germany has 1151 cases, no deaths, and only 18 recoveries (6% of resolved would only be one death).
Norway, 176 cases, no deaths, and 0 recoveries.
(3/25 3084 14 6 0.453%, but only 0.6% of cases resolved)
Sweden with 203 cases, no deaths, and 0 recoveries.
(3/25 2526 62 16 2.454%, but only 3% resolved)

Starting with zeros may seem obviously a dismissible problem now, but at the time it was a fairly common argument.

The short of it, is that the case fatality rate DURING the outbreak isn’t responding the way those touting it have assumed.

Here are links to a couple of sources explaining why it is methodologically problematic:

https://www.worldometers.info/coronavirus/coronavirus-death-rate/#correct
“One issue can be that of determining whether there is enough data to estimate T with any precision, but it is certainly not T = 0 (what is implicitly used when applying the formula current deaths / current cases to determine CFR during an ongoing outbreak).”

And

https://academic.oup.com/aje/article/162/5/479/82647

Methods for Estimating the Case Fatality Ratio for a Novel, Emerging Infectious Disease
A. C. Ghani, C. A. Donnelly, D. R. Cox, J. T. Griffin, C. Fraser, T. H. Lam, L. M. Ho, W. S. Chan, R. M. Anderson, A. J. Hedley ... Show more
American Journal of Epidemiology, Volume 162, Issue 5, 1 September 2005, Pages 479–486, https://doi.org/10.1093/aje/kwi230

This one mentioned how using the artificially low fatalities/case calculation during an outbreak can cause alarm when it starts to rise (as the situation actually improves).