Source: Analysis of new COVID-19 case counts through 7/5/2020, from the Virginia Department of Health.
The last few days have seen a possible uptick in COVID-19 cases in Virginia, as well as cases in Fairfax County. The seven-day moving average for Virginia (blue) is no longer flat, but has a slight upward slope. The seven-day moving average for Fairfax is no longer declining.
You can’t fully discount the possibility that this is just random noise. But I decided to drill down into the numbers, to the extent possible with publicly-available data, to see if there was any hint as to why.
Arguably the most interesting finding here is that the recent uptick in Fairfax cases contains a disproportionate number of school age (age 10 to 19) individuals. Not clear why that would be, though it is possible that the end of the school year resulted in more socializing within this age group.
1: Upward trend in Virginia is due to “the rest of Virginia”, i.e., the portion of the state outside of the late-reopening areas (NoVa + Richmond + Accomack County).
Source: Same as above.
For the late-reopening areas as a whole, new case counts held steady at around 200 new cases per day (blue line). By contrast, the moving average of cases for the rest of the state has risen steadily starting somewhere around 6/19/2020 or so?
I can’t plausibly link that to the 6/5/2020 start of Phase II re-opening. For one thing, the change is extremely gradual. For another, the same Phase II re-opening in NoV+ areas — about 10 days later — seems to have had no such effect.
2: Upward trend in in “the rest of Virginia” is concentrated in the Hampton Roads area, among young adults.
If I sort the counties by contribution to the recent uptick in growth in the early-reopening portion of Virginia, the Hampton Roads area comes to the top. By newspaper report, that was attributed to an increase in cases among young adults. Here’s the data for the past two weeks, in Norfolk, which is the single largest contributor to the growth in new cases. And, sure enough, the primary contributor in Norfolk is persons aged 20-29.
So this suggests that young adults, socializing, is the issue, as it has been in the currently spiking states (FL, TX, AZ and so on.) With the large concentration of naval and other military facilities there, this makes me wonder whether these cases could be military-related, or whether cases among military personnel are reported through some separate system.
In any cases, young-people-in-bars seems to be the quickest way to characterize this. Though I have no direct evidence that bars are, in fact, the issue. See Fairfax County, below.
3. The latest upward blip in Fairfax has an unusual proportion of school age kids.
The top graph shows the age mix of COVID-19 cases in Fairfax County, to date. It peaks in the working-age population. Note that teenagers (10-19, second bar) make up just over 5 percent.
But for this most recent uptick, the 10-19 year age bracket accounts for 20 percent of all new cases. Case counts are small enough that this might just be a statistical fluke. But to me, the timing suggests maybe schoolkids socializing, now that the school year is over. (The last day of school was June 12). It could also be exposure through summer jobs.
4: Fairfax cases, and the recent uptick, remain concentrated in the Latino community.
To date, just about half the COVID-19 cases in Fairfax County were for individuals who self-identified as Latino. In the past five days, that rose to about 70% (but there was also apparently a mass correction of race/ethnicity data on these records that may have affected that.
In any event, both existing cases and new cases remain strongly concentrated in the Latino community. See Post #719.