Post #1082: William and Mary update, still waiting for the other shoe to drop

Posted on March 30, 2021

Yesterday’s update of the William and Mary COVID-19 dashboard produced a few surprises.  To cut to the chase:

  • They still need to report test results for about 1300 more students.  So what’s up on the dashboard now is not the final total.
  • On Monday, they added the results of a few hundred tests with a very high positivity rate.  That suggests to me that this last batch was probably tests for students who were identified via contact tracing, or who came forward for testing because they had symptoms.
  • At present, it looks like about 2 percent of W&M students were infected with COVID during this outbreak.  That will probably fall somewhat as the last 1300 “mass testing” results are reported.

The bottom line is that this is a larger outbreak than I would have guessed, based on the data report as of last Friday.  That doesn’t mean it’s getting worse.  It just means that I under-estimated it, based on public information as of last Friday.

Details follow.

Recall what I’m doing with the W&M COVID-19 dashboard results.  I track it day by day, so that I can look at the changes from day to day.  That tells me what they added to their totals with each daily update of the dashboard.

Here’s the resulting table as of yesterday’s (3/29/2021) update.  This is, in effect, a series of snapshots of the W&M COVID-19 dashboard, strung together.  The columns at the right show the day-to-day changes.

As of 3/26/2021, W&M has reported mass testing results for just under 5000 students.  Because there are almost 6600 students currently on-campus, I expected to see the remaining 1600 tests show up yesterday.  But that didn’t happen.  (Best guess, their contract lab processes the mass-testing results on a Monday-to-Friday (non-emergency) basis.)

Instead, as you can see in yellow above, just 280 new tests showed up on Monday.  And 12.5% of those were positives.  That high positivity rate suggests to me that those tests were either from students who had symptoms, or from contact tracing of known infected students (e.g., roommates of known cases).  I can’t prove it but I think it makes sense.

All told, that last batch raises the overall positivity rate from this outbreak to 2 percent, as shown in blue above.  So, of students tested so far during this outbreak, about 2 percent have been positive. 

BUT:  That’s now a combination of mass testing (which tends to have low positivity), and a batch of symptom-driven or contact-tracing tests (which tend to have high positivity).

By my count, there are still about 1300 test results that have not yet been reported (in orange, above).  My guess is that overall 2% will drop, somewhat, when the final 1300 mass-testing test results are reported.

That said, the bottom line is that this outbreak is larger than I would have guessed, based on the results reported through last Friday.

Is there significant continued spread?

Bottom line here:  I can’t tell.  Might or might not be.

Mistakes were made on St. Patrick’s Day, and now W&M has an outbreak.  The issue now is one of containing it.  Is this outbreak still spreading, or is it in the process of being tamped down?

With that as background, it’s troubling to get a large batch of positive tests this long after St. Patrick’s Day.  Whether or not these provide any evidence of continued community spread depends on what those test were for.

If these tests arose from contact tracing — e.g., from testing roommates of persons known to be infected — then the fact that they are showing up this late has no implications for evidence of continued community spread at W&M.  These could easily be asymptomatic cases that were only tested, some days after St. Patrick’s Day, due to a known association with person who had already tested positive.

Yes, that’s continues spread of COVID-19, but it’s known and identifiable spread.  In effect, this is the part of the pandemic that is controllable.  And if that’s all this is — if this is just the fallout from diligent contact tracing by W&M — then I don’t think it’s much to worry about.

On the other hand, if these are students coming forward out of the community (i.e., without known association to existing cases), due to symptoms appearing now, then that would be a different thing entirely.

Let me work through the timing issues under that scenario.  Because if this last batch is from students spontaneously coming in for testing, that means that they were symptomatic.

The median (typical) incubation period for COVID-19 — the time between infection and onset of symptoms — is four to five days.  So, most individuals  who were going to have symptoms, from an infection directly contracted on St. Patrick’s Day, would have had them by 3/22/2021.

(Separately, a significant fraction of cases are asymptomatic or have very mild symptoms.  Estimates of that exact fraction vary widely, but 30 percent is unlikely to overstate it, particularly for a younger population.  Dr. Fauci is reported to have stated that 40 percent of U.S. cases are asymptomatic.)

The lag between testing and reporting of results is not completely clear, but it should be, at most, about three days.  My understanding is that typical Virginia lab turnaround time is 1 to 2 days, for example, as advertised by CVS here in Virginia. At some point, William and Mary guaranteed a three-day turnaround time for on-campus voluntary testing.

Based on a three-day testing lag, most of the symptomatic infections that were literally contracted on St. Patrick’s Day should have shown up on the W&M COVID-19 dashboard on or around 3/25/2021.

There’s some slack there, but on the whole, it’s getting to be a bit late for St. Patrick’s Day symptomatic infections to be showing up in the data.  So IF this last batch were symptomatic students coming forward for testing (and NOT students identified via contact tracing), then odds are that this last batch had some people who got this, after St. Patrick’s day, via community spread.  (N.B. “Community spread” is a term-of-art meaning that they don’t know where they picked up the infection.  It simply spread “in the community.” A lot of community spread would indicated an uncontrolled epidemic.)

Please note the big IF above.  This is all speculation.  Presumably W&M administration knows what fraction of this most recent batch was from contact tracing, versus symptomatic students coming forward to ask for testing.

So far, W&M is convinced that spread has mainly occurred in off-campus settings, and so the College has taken no extraordinary steps on-campus.  From what I understand of studies of spread at schools, that’s likely to be correct.  E.g., in-classroom spread is quite rare.  (I do wonder about dining halls and such.)  But so far, nobody at W&M has seen any need to take significant, lock-down-like steps to contain this.  And from that I think I can infer that if there is ongoing transmission, it’s not on-campus transmission.