06-19-2020, 01:42 AM
On Thursday June 18, Santa Clara County reported one of its 6 highest set of new cases in a day, 79.
When you look at the graph of new cases, it is appears that SCC is having a second wave. BUT, if you look closer, the number of tests has jumped somewhat in line with the apparent peaks in new cases.
17 of the last roughly 500 cases (May 30-Jun 16) have been at LTCF. Those cases aren't causing any second wave.
If you look at the percent positive cases, that has climbed but hadn't climbed too much. The nadir of percent positive was just over 1% on May 15 (week average). It climbed to 1.6% about a week later, and then ebbed to just over 1% again a week after that (May 31). My recollection is that the May 22 peak was associated with the county reaching out to test populations that were more at risk. (Possibly the tests were picking up new & not-quite-new infections. Once testing had been available for a while, there were fewer not-quite-new infections being tested.)
The SCC web site shows the cases by date of specimen collection. The daily numbers beginning June 1 have been
38, 22, 46, 36, 34, 14, 18, 41, 46, 59, 49, 47, 15, 5, 26, 20, 1 (clearly the last few days are incomplete). Presumably the 79 are spread across those days.
Positive test results by date received. Since June 1:
14, 26, 32, 28, 28, 35, 37, 28, 36, 46, 43, 44, 24, 35, 31, 40, 21
The daily percent positive numbers have been (starting Monday, May 25, 1 week earlier)
(May 25) 0.75%, 1.13%, 1.26%, 0.72%, 1.37%, 1.41%, 0.87%,
(June 1) 0.78%, 1.57%, 1.57%, 1.45%, 1.33%, 1.94%, 1.78%,
(June 8) 1.99%, 1.59%, 1.97%, 1.63%, 1.35%, 0.79%, 1.28%,
(June 15) 2.05%, 1.81%, 1.94%
I see a peak June 6 - 10 and another possible one starting June 15. (Incomplete data? June 15 & 16 look close to the expected number of tests on that day of the week. June 17 is 1/3 of the number, and is clearly incomplete results)
One theory might be that the June 15-17 numbers are secondary infections from the June 6-10 infections. Or maybe, they are both infections spread from events spaced about a week apart. (Or, there was a June 6-8 spike, with June 10-11 being its secondary infections. The serial interval is sometimes estimated to be 4 days.)
To work back to date of infection, subtract 1.64 days (SCC testing average turnaround time). Then subtract about 5 to 6 days incubation and (my guess) 1 day (maybe 2) to get tested. These are all ball park numbers. Call it 8 to 9 days. That puts the June 6-10 increase in percent positive as being around the May 29-30 demonstrations.
What is the magnitude of those peaks?
Between June 6-10: 1.84% positive % in 9901 tests
1 week earlier, May 30-June 3: 1.25% positive in 9413 tests. (Memorial Day Weekend (May 23-25) + 8-9 days is May 31-June 3, so this percent positive is probably a bit high because of that.)
2 weeks earlier, May 23-May 27, the average percent positive was 1.04%.
So, back of the envelope calculation using a baseline of 1.25%, 0.59% * 9901 tests = 58 extra cases in those 5 days. (Using 1.04% as a baseline would give 79 extra cases)
If you take June 6 to 16, and compare it to a baseline 1.3% positive percent rate,
there were an increase of about 70 cases. (baseline 1.04% would indicate 137 extra cases)
However one gets this disease, it is a potential tragedy. But I'm hopeful that the recent jump in percent positive tests is a result of a short term event rather than due to increased infection rates due to opening up in a community that is mostly trying to follow the guidelines.
When you look at the graph of new cases, it is appears that SCC is having a second wave. BUT, if you look closer, the number of tests has jumped somewhat in line with the apparent peaks in new cases.
17 of the last roughly 500 cases (May 30-Jun 16) have been at LTCF. Those cases aren't causing any second wave.
If you look at the percent positive cases, that has climbed but hadn't climbed too much. The nadir of percent positive was just over 1% on May 15 (week average). It climbed to 1.6% about a week later, and then ebbed to just over 1% again a week after that (May 31). My recollection is that the May 22 peak was associated with the county reaching out to test populations that were more at risk. (Possibly the tests were picking up new & not-quite-new infections. Once testing had been available for a while, there were fewer not-quite-new infections being tested.)
The SCC web site shows the cases by date of specimen collection. The daily numbers beginning June 1 have been
38, 22, 46, 36, 34, 14, 18, 41, 46, 59, 49, 47, 15, 5, 26, 20, 1 (clearly the last few days are incomplete). Presumably the 79 are spread across those days.
Positive test results by date received. Since June 1:
14, 26, 32, 28, 28, 35, 37, 28, 36, 46, 43, 44, 24, 35, 31, 40, 21
The daily percent positive numbers have been (starting Monday, May 25, 1 week earlier)
(May 25) 0.75%, 1.13%, 1.26%, 0.72%, 1.37%, 1.41%, 0.87%,
(June 1) 0.78%, 1.57%, 1.57%, 1.45%, 1.33%, 1.94%, 1.78%,
(June 8) 1.99%, 1.59%, 1.97%, 1.63%, 1.35%, 0.79%, 1.28%,
(June 15) 2.05%, 1.81%, 1.94%
I see a peak June 6 - 10 and another possible one starting June 15. (Incomplete data? June 15 & 16 look close to the expected number of tests on that day of the week. June 17 is 1/3 of the number, and is clearly incomplete results)
One theory might be that the June 15-17 numbers are secondary infections from the June 6-10 infections. Or maybe, they are both infections spread from events spaced about a week apart. (Or, there was a June 6-8 spike, with June 10-11 being its secondary infections. The serial interval is sometimes estimated to be 4 days.)
To work back to date of infection, subtract 1.64 days (SCC testing average turnaround time). Then subtract about 5 to 6 days incubation and (my guess) 1 day (maybe 2) to get tested. These are all ball park numbers. Call it 8 to 9 days. That puts the June 6-10 increase in percent positive as being around the May 29-30 demonstrations.
What is the magnitude of those peaks?
Between June 6-10: 1.84% positive % in 9901 tests
1 week earlier, May 30-June 3: 1.25% positive in 9413 tests. (Memorial Day Weekend (May 23-25) + 8-9 days is May 31-June 3, so this percent positive is probably a bit high because of that.)
2 weeks earlier, May 23-May 27, the average percent positive was 1.04%.
So, back of the envelope calculation using a baseline of 1.25%, 0.59% * 9901 tests = 58 extra cases in those 5 days. (Using 1.04% as a baseline would give 79 extra cases)
If you take June 6 to 16, and compare it to a baseline 1.3% positive percent rate,
there were an increase of about 70 cases. (baseline 1.04% would indicate 137 extra cases)
However one gets this disease, it is a potential tragedy. But I'm hopeful that the recent jump in percent positive tests is a result of a short term event rather than due to increased infection rates due to opening up in a community that is mostly trying to follow the guidelines.