I'm seeing a lot of people uncritically retweeting this report that implies China hid ~40k COVID deaths in Wuhan.
If there's reason to be wary of China's official figures (and there is!), there's also ample reason to as/more wary of this number.
🧵
If there's reason to be wary of China's official figures (and there is!), there's also ample reason to as/more wary of this number.
🧵
First, even if China hid or under-counted a significant number of deaths, 40K (vs official figure of ~2500) would be wildly implausible.
As @kakape lays out in this thread, fatality rates we've seen for this virus range from <1% to China's ~4%.
As @kakape lays out in this thread, fatality rates we've seen for this virus range from <1% to China's ~4%.
So 40k deaths at <1% CFR would mean >4m people in Wuhan had been infected (of total pop'n of 11m).
If 4% CFR, then still 1m infected.
Either would require a MASSIVELY more aggressive level of transmissibility than we've seen anywhere from this virus. Not plausible.
If 4% CFR, then still 1m infected.
Either would require a MASSIVELY more aggressive level of transmissibility than we've seen anywhere from this virus. Not plausible.
In fact transmissibility elsewhere has been pretty consistent with China's official figures, as this @ECDC_EU report on Italy finds (R0 2.76 - 3.25).
This holds with Seattle as well, which had an initial R0 of 2.7 prior to social distancing.
ecdc.europa.eu/sites/default/…
This holds with Seattle as well, which had an initial R0 of 2.7 prior to social distancing.
ecdc.europa.eu/sites/default/…
So the 40k figure would be wildly out of step with what we know of the characteristics of this virus.
There's a reason for that. The ~40k figure is extrapolated entirely from anecdotal assertions that the city's 7 crematoriums "have been working around the clock."
There's a reason for that. The ~40k figure is extrapolated entirely from anecdotal assertions that the city's 7 crematoriums "have been working around the clock."
The 40k+ estimates are base on how many deceased people *could* hypothetically be cremated (from all causes!) if all 7 crematoriums were working non-stop.
That's quite different from affirmative evidence that 40k people died specifically of COVID.
That's quite different from affirmative evidence that 40k people died specifically of COVID.
The article also implies the crematoriums are clearing a backlog of deaths, but presumably those would not all be COVID (or even mostly COVID, since people continue dying of other normal things during an outbreak, and perhaps at elevated levels due to strained hospital capacity).
So from an epi perspective, a 40k figures doesn't hold up.
And as a methodology for measuring COVID mortality, extrapolating from hypothetical max cremation capacity doesn't hold up well either.
And as a methodology for measuring COVID mortality, extrapolating from hypothetical max cremation capacity doesn't hold up well either.
Did more people die in Wuhan than reported over the past three months?
Quite probably. Some %age of COVID deaths were likely missed in the count, willfully or not. And mortality from other causes would likely be elevated as well due to strained hospitals.
Quite probably. Some %age of COVID deaths were likely missed in the count, willfully or not. And mortality from other causes would likely be elevated as well due to strained hospitals.
There are a lot of people out there motivated to find any way to make US death projections look less bad.
But leaping from this article to "China is hiding 40k COVID deaths", as I'm seeing in my feed, doesn't hold up.
But leaping from this article to "China is hiding 40k COVID deaths", as I'm seeing in my feed, doesn't hold up.
I say this not to defend China's handling of this outbreak. They made many mistakes that made this worse than needed - mostly for their own people.
But there's a lot of goalpost-moving going on in the US and elsewhere now given how badly Western countries have handled this.
But there's a lot of goalpost-moving going on in the US and elsewhere now given how badly Western countries have handled this.
So we should be subjecting anyone's numbers to comparable scrutiny, rather than cherry-picking reports that reinforce a preferred political narrative.
More on this:
The key info needed to assess the high risk of COVID was made available via @WHO by 23 January:
- Novel respiratory coronavirus
- Efficient human-to-human transmission
- R0 of ~2+
- Fatality rate of >1%
- Severe cases ~25%
who.int/news-room/deta…
The key info needed to assess the high risk of COVID was made available via @WHO by 23 January:
- Novel respiratory coronavirus
- Efficient human-to-human transmission
- R0 of ~2+
- Fatality rate of >1%
- Severe cases ~25%
who.int/news-room/deta…
These data have been refined through further research since, but were all ballpark correct.
The question of asymptomatic/presymptomatic transmission was not fully settled at that point but it was clearly a known possibility and we being reported by Chinese docs.
The question of asymptomatic/presymptomatic transmission was not fully settled at that point but it was clearly a known possibility and we being reported by Chinese docs.
All the "China withheld info yada yada" overlooks (willfully ignores) that by the end of January - two months ago! - we had all the info we needed to be deeply concerned and to trigger preparedness.
We just didn't do so.
We just didn't do so.
One last thought - China will reportedly initiate serologic testing (which can confirm whether someone previously had the virus). If done systematically and shared openly, this would reveal the actual level of infections among the population. statnews.com/2020/03/11/cdc…
