Let’s talk about a COVID19 seroprevalence study, this time well conducted (it seems), which provides a petri-dish type natural experiment. It’s the small Austrian mountain ski village of Ischgl, one of the epicenters of SARSCoV2. Thanks to @florianederer for letting me know! 1/
In this Austrian village:
- 1,473 tested (79% of population) in late April, 2020
- 42.4% have SARSCoV2 antibodies (N=625)
- 85% asymptomatic (N=94 symptomatic cases)
- 9 hospitalizations
- 2 deaths
2/
- 1,473 tested (79% of population) in late April, 2020
- 42.4% have SARSCoV2 antibodies (N=625)
- 85% asymptomatic (N=94 symptomatic cases)
- 9 hospitalizations
- 2 deaths
2/
With some calculation, this yields:
IFR = 0.32% (which is very typical of SARSCoV2)
And, given low level of symptomatic cases (perhaps due to young age?),
sCFR = 2.1% (which is very high!) (2/94)
(Most studies show about ~50% asymptomatic cases of COVID19.)
3/
IFR = 0.32% (which is very typical of SARSCoV2)
And, given low level of symptomatic cases (perhaps due to young age?),
sCFR = 2.1% (which is very high!) (2/94)
(Most studies show about ~50% asymptomatic cases of COVID19.)
3/
I would guess age in this village skews young, which would explain low symptomatic case percentage and low-end IFR. Here is the only info yet released:
- 1,259 adults
- 214 children (<18)
- 479 households
4/
- 1,259 adults
- 214 children (<18)
- 479 households
4/
As is typical of SARSCoV2, the attack rate is lower in the young in this village: only 27% of tested children have antibodies (N=58). This means there were 625-58=567 adult cases. Which means attack rate in adults is 567/1,259=45.0%.
5/
5/
This overall attack rate is still a bit low for SARSCoV2, especially since well-mixed assumption probably holds in a small village like this. Herd immunity is reached (subject to assumptions) at (R0 - 1)/R0, which is 61.5% for R0=2.6 (which is what we estimate for SARSCoV2). 6/
But, conversely, if epidemic is over in this village, which seems likely, calculation with above data yields an R0 of 1.9, which is on low end of prior outbreaks for SARSCoV2. But this would be good news from a transmissibility point of view, with respect to this pathogen. 7/
The results have only been released via press release so far (in German): tirol.orf.at/stories/305482… My thanks to @florianederer for bringing these fascinating data to my attention. NB @florian_krammer too. 8/
PS The press release hilariously mentions that the low rate of hospitalizations is due to the tough mountain spirit in this Austrian village, I am told. 9/
IFR = Infection Fatality Rate (fraction of people who get infected who die, whether they get symptoms or not)
CFR = Case Fatality Rate (fraction of infected people who come to medical attention who die)
sCFR = Symptomatic CFR (fraction of people with symptoms who die)
10/
CFR = Case Fatality Rate (fraction of infected people who come to medical attention who die)
sCFR = Symptomatic CFR (fraction of people with symptoms who die)
10/
Here is the original press release, with better and clearer detail about some of the above numbers i-med.ac.at/mypoint/news/7… (again, thanks to @florianederer). 11/
There is ambiguity in report that 85% of people in this village were "asymptomatic," & (as I had noted) that was suspiciously high. So sCFR in this village is likely similar to other studies (good!). Must await full study. IFR, attack rate, & R0 estimate unchanged by this. 12/
Here is an upcoming event that may provide more information about the case of the Austrian mountain village that may have achieved herd immunity with COVID-19 (via @florian_krammer) NB @florianederer 13/
