Post #758: Data drill-down for Norfolk and Virginia Beach

It has become a young person’s pandemic in parts of Virginia.

Source:  Analysis of counts COVID-19 cases from the Virginia Department of Health.


Where are the new COVID-19 cases appearing in Virginia?

This past week’s uptick in new COVID-19 cases in Virginia has been concentrated in the Hampton Roads area (see Post #757).  Yesterday, two adjacent cities at the mouth of the Chesapeake Bay — Virginia Beach and Norfolk — accounted for 25% of all new Virginia cases.  But they account for just 8% of Virginia’s population.  The data are shown in the table below.

It’s not that these counties have had a lot of cases so far, compared to other parts of the state.  So far, Northern Virginia was hit far harder by COVID-19 than the Hampton Roads area was.  As of today, 1.3 percent of Fairfax County residents have been diagnosed with COVID-19.  None of these areas exceeds 1 percent in the “cumulative % of pop …”.  So it’s not the total cases to date that’s the issue.

The problem is how rapidly those case counts are now growing, and that the growth is accelerating.  For Virginia Beach and Norfolk, you can see that yesterday’s increase was well above the average for the week.

And if you play out what would happen if yesterday’s rate of new case counts were to continue for a year, you get the right-hand column of the table above.  That’s the  “annualized per-capita infection rate”.  That’s the fraction of the population that would be newly infected with COVID-19 if yesterday’s new infection rate held steady for a year.

To put Norfolk’s rate into perspective, here’s a snapshot of a few areas, and their annualized per-capita infection rate yesterday:

Yesterday’s (7/15/2020) new cases per capita, annualized, in various places:

20.4% Norfolk, VA
16.5% Florida
16.3% Arizona
11.7% Virginia Beach, VA
9.9% Texas
4.6% Virginia

For Norfolk, VA, that’s only one day, and that’s only one city, but that’s still in the zone where the rate of new infections is (or should be) ringing alarm bells somewhere.  Even though they don’t have a lot of infections per capita right now


Drill down.

What did yesterday’s new cases in Norfolk and Virginia Beach look like?  The data released by the Commonwealth allow me to “drill down” three ways, one at a time:  Age, race/ethnicity, and ZIP code.  To get a reasonable sample size, let me take all the new cases in the last two days.

New cases in those two localities have a much large portion of young adults, those age 20 to 29.  Where the comparison group is all cases, to date, in Virginia.  Unfortunately, this new focus on young adults is something they have in common with the most out-of-control states (AZ, FL).

I have guessed that this situation — a lot of diagnosed cases among young adults — naturally leads to unexpectedly rapid growth going forward (Post #755).  Read that if you want an explanation.   But the bottom line is that, for disease spread, cases diagnosed among the young are likely far more important, because, that low age indicates that there is a much larger pool of undiagnosed cases is circulating in the population than you would guess from historical norms.

Race/ethnicity.  It’s not clear that the race/ethnicity data are useful because in both cities, the single largest category is “not reported”.   For what it’s worth, if I only look at cases where race/ethnicity was reported, then:

About two-thirds of total cases and three-quarters of new cases in Norfolk are among African-Americans.  That compares to about 41% of the population there being non-hispanic African-Americans (reference).  So the COVID-19 burden is somewhat disproportionate, but not hugely so (relative to, say, what was observed in the Latino population in the Northern Virginia area, Post #719).  In Virginia Beach, by contrast, about half of all cases and 60% of new cases were among the non-Hispanic white population.  Versus about 60% of the population there being non-Hispanic white.

By ZIP code, I’m not familiar enough with the area to make much of the map.  Here, bigger dots = higher levels of new cases per capita yesterday.  Norfolk is to the left, Virginia Beach to the right.  The Atlantic ocean is to the right.

Surprisingly, there wasn’t a very strong link between income and new cases per capita.  Possibly, that’s because I’m only looking at a day’s worth of data.  That said, the worst-one-third and middle-one-third of zip codes, in terms of yesterday’s increase, had virtually identical average IRS incomes.  Only the very top third of ZIPs had modestly higher average income.

Make of that what you will.  When I did that analysis for Northern Virginia, it looked like this was a working-person’s pandemic, and income mattered pretty strongly.  There was a nice, almost straight-line relationship, between prevalence of disease and income.  Here, this is a young person’s pandemic, and it seems to be circulating among the young, to some degree regardless of area income.  That’s a weak indication that this is as it has been classified in other areas — it’s young people gathering socially, instead of public-facing workers picking this up at work.

All in all, what’s happening now in the Norfolk/Virginia Beach area seems to match the stylized facts of the rapid growth in AZ, FL, TX and other places.  It’s a young person’s pandemic, and it’s only marginally linked to either race or to income.