4 ms·

Interesting metrics. Indicator 1 (cases) and 2 (hospitalization) are clearly met. Indicator 3 (test capacity) has a very aggressive target. 200 tests a day pe

Interesting metrics. Indicator 1 (cases) and 2 (hospitalization) are clearly met. Indicator 3 (test capacity) has a very aggressive target. 200 tests a day per 100k? Why must it be this high? (Germany by comparison, opening schools now, is at 80 per 100k). Santa Clara County is already at a 3% positive rate week to date (Korea level during their Feb epidemic) Indicator 4 (case investigation) status is not disclosed but I would guess is met or close to met in Santa Clara county given how steep the drop-off has been in cases and confirmation by other counties (Contra Costa). Of course, it gets harder to contact trace as you ease restrictions. Indicator 5 is a bit fuzzy - supply under what hospitalization assumptions?

contemporary343 7y ago
Interestingly, LA county is nearly at that target. The public health system (excluding internal test capabilities at hospitals etc) can do 16K tests/ day for a population of 10 million. So about 160 tests a day/ 100 K.
SpicyLemonZest 7y ago
The Bay Area counties included here are all pretty dense, so it makes sense that they'd want to have more test coverage per capita than a full country might need.
valuearb 7y ago
3) might be linked to cases per million and relative growth rates. Germany has roughly 400 active cases per million, California has about 1,000, and I wouldn’t be surprised if the Bay Area is well above CA as a whole given its density.
- usaar333 7y ago
  Bay Area is well below CA as a whole. Cases have grown by ~375/million in last 14 days to proxy for "active" cases. Santa Clara County's growth is only ~235/million
cataflam 7y ago
If it can give you a comparison point, France wants to be able to do 700k tests per week in less than 2 weeks time to have test and trace capability and start lifting the lockdown. That comes out to approximately 150 tests a day per 100k. The following reason was given. The models predict 1000-3000 new cases a day around the time lockdown would be lifted. Each new case will imply 20-25 other persons in contact in the preceding days. 3000 x 25 = 75k tests per day, ie ~115 per day per 100k. Going to 150 allow an extra margin to also set up testing campaigns in sensitive places, eg nursing homes.
wpietri 7y ago
I suspect 5 isn't fuzzy in practice. My guess is that they delegate the details to each hospital, which will have very detailed notions of what they need. Especially given that their supply people will have been sweating this question for the last six weeks.
hardtke 7y ago
Regarding #4, with the ubiquity of call blocking, I don't see how it is possible for them to reach 90% of the people that may have been exposed, especially once people find out that picking up that phone call means you will be in quarantine. To make matters worse, the authorities likely won't be able to tell you where/when you were exposed and to whom (medical privacy). This target is basically unreachable IMHO.
- mindslight 7y ago
  There are more ways to contact people than just calling them. Your comment stinks of why we can't get anything done these days - it starts with an assumption of failure, and argues for giving up based on it.
- SpicyLemonZest 7y ago
  A round number repeated for 4 metrics in a row is never really meant to be taken seriously. If it turns out that they can only reach 80% of contacts, they'll adjust the plan to account for that - nobody's going to let it be the one factor that keeps restrictions from being loosened.
- stallmanite 7y ago
  I guess this is another reason something should have been done about the spam calls. Who answers the phone now for an unknown number?
rtkwe 7y ago
> Indicator 3 (test capacity) has a very aggressive target. 200 tests a day per 100k? Why must it be this high? (Germany by comparison, opening schools now, is at 80 per 100k). Santa Clara County is already at a 3% positive rate week to date (Korea level during their Feb epidemic) To actually be able to control the outbreak while people are out and about you need to be able to test lots of people whenever a new case is discovered to see if they actually passed it on to anyone. Each new case will trigger many cases especially if they're showing up sick at the hospital/doctor which means they've potentially been spreading the disease for 2 weeks.
abhorrence 7y ago
3% week to date might change (in either direction). SCC reports tests by date of specimen collection, so the last 2 days are definitely not complete data, and a significant portion of the last 5 days are not either.

contemporary343 7y ago

Interestingly, LA county is nearly at that target. The public health system (excluding internal test capabilities at hospitals etc) can do 16K tests/ day for a population of 10 million. So about 160 tests a day/ 100 K.

SpicyLemonZest 7y ago

The Bay Area counties included here are all pretty dense, so it makes sense that they'd want to have more test coverage per capita than a full country might need.

valuearb 7y ago

3) might be linked to cases per million and relative growth rates. Germany has roughly 400 active cases per million, California has about 1,000, and I wouldn’t be surprised if the Bay Area is well above CA as a whole given its density.

usaar333 7y ago
Bay Area is well below CA as a whole. Cases have grown by ~375/million in last 14 days to proxy for "active" cases. Santa Clara County's growth is only ~235/million

cataflam 7y ago

If it can give you a comparison point, France wants to be able to do 700k tests per week in less than 2 weeks time to have test and trace capability and start lifting the lockdown. That comes out to approximately 150 tests a day per 100k. The following reason was given. The models predict 1000-3000 new cases a day around the time lockdown would be lifted. Each new case will imply 20-25 other persons in contact in the preceding days. 3000 x 25 = 75k tests per day, ie ~115 per day per 100k. Going to 150 allow an extra margin to also set up testing campaigns in sensitive places, eg nursing homes.

wpietri 7y ago

I suspect 5 isn't fuzzy in practice. My guess is that they delegate the details to each hospital, which will have very detailed notions of what they need. Especially given that their supply people will have been sweating this question for the last six weeks.

hardtke 7y ago

Regarding #4, with the ubiquity of call blocking, I don't see how it is possible for them to reach 90% of the people that may have been exposed, especially once people find out that picking up that phone call means you will be in quarantine. To make matters worse, the authorities likely won't be able to tell you where/when you were exposed and to whom (medical privacy). This target is basically unreachable IMHO.

mindslight 7y ago
There are more ways to contact people than just calling them. Your comment stinks of why we can't get anything done these days - it starts with an assumption of failure, and argues for giving up based on it.
SpicyLemonZest 7y ago
A round number repeated for 4 metrics in a row is never really meant to be taken seriously. If it turns out that they can only reach 80% of contacts, they'll adjust the plan to account for that - nobody's going to let it be the one factor that keeps restrictions from being loosened.
stallmanite 7y ago
I guess this is another reason something should have been done about the spam calls. Who answers the phone now for an unknown number?

rtkwe 7y ago

> Indicator 3 (test capacity) has a very aggressive target. 200 tests a day per 100k? Why must it be this high? (Germany by comparison, opening schools now, is at 80 per 100k). Santa Clara County is already at a 3% positive rate week to date (Korea level during their Feb epidemic) To actually be able to control the outbreak while people are out and about you need to be able to test lots of people whenever a new case is discovered to see if they actually passed it on to anyone. Each new case will trigger many cases especially if they're showing up sick at the hospital/doctor which means they've potentially been spreading the disease for 2 weeks.

abhorrence 7y ago

3% week to date might change (in either direction). SCC reports tests by date of specimen collection, so the last 2 days are definitely not complete data, and a significant portion of the last 5 days are not either.