Using a delay-adjusted case fatality ratio to estimate under-reporting (paper not yet peer reviewed)

[muriel_volestrangler]

The lead author on this is from the London School of Hygiene and Tropical Medicine, which does have an excellent reputation; but it hasn't been peer-reviewed yet, so there may be assumptions that other experts will point out. They're updating it with new figures at least daily.

https://cmmid.github.io/topics/covid19/severity/global_cfr_estimates.html

Its current estimate is that only between 12 and 18% of cases have been reported in the USA; in some countries it's a lot more (right up to 59-100% for South Korea, with Germany not far behind), or less (4.3-5.6% for Spain, or 4.4-5.7% for Italy).

If I apply their midpoint figures to the official case numbers here, it shows the death rate (assuming all deaths have been properly reported) ranges from 0.18% (the USA, with Canada close at 0.2%) to 1.3% for China. It also implies over 2% of Italy and Spain have already got the virus.

[pat_k]

1. Thank You for this.

With the rate of testing so low in most states, it is so difficult to even "guessimate" what real numbers could possibly be. I'll be watching for updates, particularly as better data is available from the recently "ramped up" testing in WA and NY.

On the general subject of data, the number of hospital admissions with COVID-19-like symptoms over the past four weeks in WA was interesting to me. I am probably reading too much into it, but it seems the closures at the beginning of Mar may already be having an effect on "leveling off" infection rates. (Of course, these are just "COVID-19-like symptoms -- doesn't include numbers of cases that were later confirmed to be, no not to be, COVID-19.)

From this OP:

02/23-02/29 61
03/01-03/07 126
03/08-03/14 229
03/15-03/21 254