If we make the probably bad assumption that those tested, other than for having symptoms, in Santa Clara County are representative of the population, and we recognized that because this includes those with symptoms, this is an upper bound, then I think a reasonable upper bound estimate of current infections in Santa Clara County is the positive test percentage in the county for a week taken 5 days ago. (5 days ago to allow for all tests to be returned; a week covers the weekend drops).
How much an upper bound this is, I can't really say. For all I know it might be 10x the real rate.
By those criteria (average positive % testing, week ending May 13), in Santa Clara County, there are (upper bound) 1.22% that would test positive on any particular day (1.25% if averaged over 2 weeks, so the infection rate is very slowly dropping).
For a population of 2 Million, that's about (upper bound) 25K PCR-detectable infections on any particular day.
I don't have a solid number for how many days is typical for an infection to be detected by a PCR test, so I'll use 3 weeks (shedding duration is clearly shorter; my recollection is of an 18-day average detection but I don't remember if that was from infection or onset of symptoms). That corresponds to 1190 new infections per week, or about 170 per day. For that week, SCC had 103 positive tests. So roughly, total new infections is (upper bound) 12x those with positive tests.
Somewhat separably, of those 25K people that would be PCR-positive on any day, using about a 10-day infectious window and the 3-week PCR-detectable window, about (upper bound) 12K would be infectious on any particular day. That's (upper bound) 6 out of every 1000 people.
This is across the county. Different neighborhoods & different situations (particularly, congregate housing) result in different percentages.
Lots of assumptions there. Make your peer review & feel free to make your own estimate. Do note this is an estimate of those currently infected (not ever-infected as an antibody survey might detect).
By comparison, L.A. county (minus Pasadena & Long Beach) numbers for May 13 are 7.67% (1 week) and 5.84% (2 week). L.A. county had two weeks of high testing numbers that drove their average down, but the past week has been about 60% as much testing. (I don't know if it less capacity, less demand, or delayed results.).
How much an upper bound this is, I can't really say. For all I know it might be 10x the real rate.
By those criteria (average positive % testing, week ending May 13), in Santa Clara County, there are (upper bound) 1.22% that would test positive on any particular day (1.25% if averaged over 2 weeks, so the infection rate is very slowly dropping).
For a population of 2 Million, that's about (upper bound) 25K PCR-detectable infections on any particular day.
I don't have a solid number for how many days is typical for an infection to be detected by a PCR test, so I'll use 3 weeks (shedding duration is clearly shorter; my recollection is of an 18-day average detection but I don't remember if that was from infection or onset of symptoms). That corresponds to 1190 new infections per week, or about 170 per day. For that week, SCC had 103 positive tests. So roughly, total new infections is (upper bound) 12x those with positive tests.
Somewhat separably, of those 25K people that would be PCR-positive on any day, using about a 10-day infectious window and the 3-week PCR-detectable window, about (upper bound) 12K would be infectious on any particular day. That's (upper bound) 6 out of every 1000 people.
This is across the county. Different neighborhoods & different situations (particularly, congregate housing) result in different percentages.
Lots of assumptions there. Make your peer review & feel free to make your own estimate. Do note this is an estimate of those currently infected (not ever-infected as an antibody survey might detect).
By comparison, L.A. county (minus Pasadena & Long Beach) numbers for May 13 are 7.67% (1 week) and 5.84% (2 week). L.A. county had two weeks of high testing numbers that drove their average down, but the past week has been about 60% as much testing. (I don't know if it less capacity, less demand, or delayed results.).