Are we not taking Covid-19 seriously

[DG_player]

Here's an interesting tool to assess your disc golf tournament or thanksgiving dinner risk:

https://covid19risk.biosci.gatech.edu

[DavidSauls]

Interesting, though doesn't that assume a representative sample of the population at your event?

[thirtydirtybirds]

It appears to, but it does break down by county, so it looks to be a very good tool for risk in a specific area. It’s at least a very strong baseline.

[DG_player]

Originally Posted by DavidSauls

Interesting, though doesn't that assume a representative sample of the population at your event?

Yes, it's also making some broad assumptions of prevalence vs. confirmed cases. I think it's more useful to just give some perspective on how high the risks actually are. I was a little shocked what it estimated for my area, given I'm in one of the better doing counties in the country.

[DiscFifty]

Originally Posted by DG_player

Here's an interesting tool to assess your disc golf tournament or thanksgiving dinner risk:

https://covid19risk.biosci.gatech.edu

"Based on seroprevalence data, we assume there are ten times more cases than are being reported (10:1 ascertainment bias)."

I tend to agree there could be 5-10x the number of actual active cases reported. Still a bit confused if that would be a good or bad thing. The death rate would be much lower for instance.

[DG_player]

Originally Posted by DiscFifty

"Based on seroprevalence data, we assume there are ten times more cases than are being reported (10:1 ascertainment bias)."

I tend to agree there could be 5-10x the number of actual active cases reported. Still a bit confused if that would be a good or bad thing. The death rate would be much lower for instance.

Higher is obviously better. At 10x, based on the amount of resolved confirmed cases in the U.S. the death rate would be around 0.35% which is much better than original estimates.

[dysmike]

Originally Posted by DG_player

Higher is obviously better. At 10x, based on the amount of resolved confirmed cases in the U.S. the death rate would be around 0.35% which is much better than original estimates.

Improved knowledge of care, and not having an over-whelmed care system (well, it's probably close at this point) helps.

[DavidSauls]

It's probably better for overall planning purposes -- which do have to take into consideration the range of the population -- than for someone weighing the risks of Thanksgiving.

I'd just seen, elsewhere this morning, that my county is 4th in positive tests in my state. It's a rural county with a small population, so the overall numbers aren't striking.

Positive tests, as a ratio of population, are 4.5%. If the true rate is 10x that, does it mean that almost half the county population has been infected?

[dysmike]

Positivity rate is (positive tests)/(total tests) x 100%

So no, because we can not assume that everyone who is positive has been tested, or even has symptoms. So the true rate of infection is much fuzzier.

[DG_player]

Originally Posted by DavidSauls

It's probably better for overall planning purposes -- which do have to take into consideration the range of the population -- than for someone weighing the risks of Thanksgiving.

I'd just seen, elsewhere this morning, that my county is 4th in positive tests in my state. It's a rural county with a small population, so the overall numbers aren't striking.

Positive tests, as a ratio of population, are 4.5%. If the true rate is 10x that, does it mean that almost half the county population has been infected?

If you're saying 4.5% of your population has tested positive to date, then yes if the 10x estimate is correct 45% are or have been infected in the past. If it's 5x then only 22.5%. I'm guessing it's pretty varied across the country though depending on when the cases happened and the accessibility of tests. The ratio was probably much high early on than it is now just given the availability of testing.