Virginia COVID-19 Status for Sept. 7: Still college outbreaks

 

Daily Status, Sept. 7

Reminder:  any sections that are unchanged since yesterday are grayed out.

I wrote a supplementary blog describing how I monitor colleges from zipcodes.

I have been looking at the COVID-19 numbers in Virginia since March 8;  I started writing this first with Facebook posts in late March, sharing them to VVCP in Mid-May, and converting to a blog format in Mid August.  That mean, I have noticed trends.  Some are statistically significant, some are not.  One trend is the Monday of a three day weekend will not be typical.  Today is the third such event.  The first was Memorial Day -- that was high because of supplementary testing done in Northern VA over the weekend.  But July 6 (Monday) remains Virginia's lowest number of reported cases since mid April.  Today, is also low -- nearly the lowest number since the July 4 weekend:  645 cases.  I do not think today's or tomorrow's numbers will be particularly meaningful because of the holiday.  Therefore, I do not plan to spend much time with the analysis

Summary/Situational Awareness


Usually, during the COVID-19 pandemic, at a given time, parts of the commonwealth have low case counts, parts have  high COVID-19 numbers...some areas are improving, and others are getting worse.  That remains the case now, with one group (colleges) increasing the COVID-19 number by 10% per day, while the state COVID-19 numbers (excluding that group) are essentially flat.  The good news is few of the hospitalizations for COVID-19 are from college students (possibly 1 from VA Tech, 3 from JMU, and 2 more throughout the state).  In terms of numbers, the schools that were having issues (UVA, VT, Radford, JMU, VCU) continue to have issues.  The other schools are doing ok.  

Across the commonwealth of Virginia, there were 645 reported new cases of COVID-19.  The number is very low but not surprising for a Monday of a long weekend. Overall, cases are increasing at just over 1% per day. It is worth noting that 12% (81) of the cases are directly related to college; however the college number is probably unrealistically low for the day.  Regionally, only SW VA & NW VA are increasing, largely the result of the outbreak at Radford, VT, and JMU.  The testing numbers across the state remain ok, but not great, at around 7% positive.  I would like to see the overall positive rate under 5%

Outside of college towns and a few smaller cities/towns, the state is doing well. In fact, 21% of the probable active COVID cases are affiliated with a the colleges of Radford, JMU, VT, UVA, and VCU; those colleges account for 1.5% of the population.



Regions:

Over the last several weeks, we have seen a trend where the more rural parts of the state, with a few isolated large colleges have been trending poorly compared the more urban setting (even if those with colleges).  For example, Hampton Roads and NOVA have not seen impacts with the colleges reopening whereas SW & NW are being swamped.  In particular, the New River area is showing 0.8% of the population positive in the last 10 days or so.  This is the result of the annual migration of college students to Blacksburg and Radford.  







Regional growth rates are (in fraction per day)


NOVA: 1.000 -- GMU

Central VA: 0.996 -- UVA & VCU

Hampton Roads/Eastern VA:  0.9966 -- W&M, CNU & ODU

SW VA: 1.013 -- VT & Radford

NW VA:  1.012-- JMU

 

The state as a whole is is increasing at 1.012


Daily case count per 100,000 is a often sited metric (below 10 (0.01%) statewide would take VA off the quarantine list .  The following table shows the number per region.


Daily Cases/100,000 
(corrected for testing)

Region

Last month

Last week

NOVA

 9.5

8.9

Eastern

12.7

 9.1

Central

12.1

11.9

NW

 8.5

16.6

SW

13.9

20.5


The following charts are for the 5 regions/trends.  Again, SW VA, even with the lower population, has more cases than any other region.










Local/Northern VA:

The overall trend in Northern Virginia indicates, that after correcting for testing, the case load has dropped by more than a factor of 10x since our early may peak, though significantly higher than the early July minimum.  However, over the last several weeks, Northern VA has been essentially flat, at about 240 new confirmed cases a day, but over the last week, has started to increase again.  The current rate is  more than twice the rate from early July (120 / day), but down from about 3000 cases per day at the peak (because of the availability of testing, 40% of the tests were positive in northern VA, indicating that only 1 in 8 cases were confirmed).  Hospitalizations and deaths are also down from the peak, but up in the last month.



Right now, no jurisdiction is increasing by more than 1%.  But, Fairfax and NOVA is increasing.  A 1% increase (1.01 Rt) is doubling in 70 days.  

Fairfax Co.: 1.004
Arlington Co.: 0.967
City of Alexandria: 0.985
Prince William Co.: 0.997
Loudoun Co.: 1.012
The growth rate is defined as Rt, or the growth time constant.  Simply, todays number are approximately the growth rate times yesterday's numbers.  This is the exponential time constant.  Fortunately, the time constants are about 1, and our rate of cases is about 7/100,000 per day.  Ideally, we would be 0, but 7 is much better than our peak in which was around 30/100000K. 


About 10 day ago, Arlington began enforcing social distancing guidelines. Since then, there trend has gone from 1% per day to -5% per day.  Now, Arlington is doing the best; Alexandria and PW are doing worse.





What is most noteworthy of the NOVA Zip code map is the the overall lack of contrast:  the bad areas are doing better, and the good areas are doing worse, with the exception of two areas near the fringe of "Northern VA".  (The state would call those NW VA). 



In Fairfax County, the cases are about static over the last few weeks.  By comparison, Arlington and Alexandria are showing significant declines since Arlington started enforcing social distancing in late August. By contrast, Vienna and Mclean, while still very good, have shown an increase.

Region

Last month

Last week

Fairfax Co

Vienna

 4.5

 7.2

McLean

 5.2

 5.3

 So. Alexandria

12.5

 9.6

 Reston/Herndon

 9.8

 6.9

 Annadale/Fall Church

12.4

11.62

 Fairfaix

 8.0

 7.5

Arlington/Alexandria

 No. Arlington

 8.3

 4.3

 So. Arlington

13.9

 8.0

 Alexandria City

14.3

13.6






Age Distribution: 

First of all, I fixed the problem with this chart.

The trends/growth for the the middle aged gas dropped to be similar to the teen/tween and elderly. If we look at state-wide numbers, we see a very different trend set:  all ages are doing well except those that include college ages (18-24).  As we can see, NOVA flollows the growth trend of the state, except for teens/tweens.  I am confident that growth is driven by students at college.

             <10    0-19.  20-39.  39-59.  60+
  NOVA  1.01    1.01    1.00      1.00   1.00
  AllVA   0.99    1.04    1.01      1.00   1.00

Unfortunately, I can not parse out 18-24 seperate from the other groups.




Colleges:

Before I get into details, there are at least three COVID hospitalizations in colleges that are from ages consistent with college students.  3 at JMU, one at VT.  There could be more, if upon admission, the student gave the home address.

Note that the reported numbers were low at all colleges yesterday.  I suspect this was related to the reporting being in the middle of the long weekend.  I am not changing anything other than the tables and charts based on today's numbers.

My process combines the VA Department of Health data and what is reported by the colleges.  


RED means there is clear evidence for community spread
YELLOW means there may be community spread; still ambigious
GREEN means no evidence of community spread
BLACK means they went online.

College

% Positive

VDH Cases*

Estimates Cases**

Dasshboard Cases

% of population infected ***

Va Tech

19.611.6 

527

2066

416*****

6.1

GMU

3.8

22

22

22

0.001

UVA

13.7

118

323

161

1.4

ODU

 4.5

0

0

16*****

 0

JMU

34.7

  627

  >2000

 999

>10%

CNU

9.5

0

0

5

0.000

UMW

5.5

0

0

N/A

0.000

Radford

88.2

483

 >1000

293*****

>10%

VCU

9.2

135

 249

183*****

0.008

W&M

1.4

0

0

0

0.000

* VDH cases are based on the cases either in the zipcodes containing the student population or the community.  In addition, I remove the rate of cases prior to students arriving on campus
** estimated number of cases is an attempt to normalize for testing limitations. Specifically, I assume at 5% positive, 100% of the cases would be caught. so I normalize it to that value.  If the % positive is very high (>40%) I am likely overestimating the numbers.
*** Dashboard cases are only counted if I can find the dashboard.  In some cases, it is difficult to distinguish positive tests from cases (1 case may have multiple positive tests; that is mostly at VT).  I include active cases if reported, otherwise, I use total cases.
****% population uses the total reported number of students rather than just those on campus; it may be off when the percent positive is above >40%.
***** Old data, not updated for today.
****** Reported in the press

Large Scale Community Spread:

RADFORD (linked to the university Dashboard)

Radford administration seems to think the virus is under control, according to press reports.  However, as long as the percent positive is running more than 50% (the state reports more than 85% of tests are positive), it tells me they are missing a lot of positive cases.  I can not assess how things are:  they are testing at full capacity; the growth rate in confirmed cases reflects the testing ability rather than the caseload. At this point, my only recommendation is to perform serology/antibody studies to retrospectively determine the number of cases.   Radford tested the students prior to arrival, but allowed students on campus prior to the test results.

JMU

JMU is going online.  JMU has been fairly transparent with the situation, but could not get ahead of it.  At this point, it is mitigation.    If you have a student from JMU, try to get them tested prior to bringing them home, and it would be advisable to assume the student is infected.  There are now two hospitalizations in Harrisonburg that are of the age group expected for College students.  JMU did not require testing prior to arrival on campus.
We have seen the prevalence of the virus in Blacksburg increase with the return of the students. The most concerning aspect of Va Tech is 17% of the tests are coming back positive.  That suggests that either they are rationing tests, or they are undercounting.  Either way, I estimate indicate more than 5% of the student body is positive. VT will start updating the dashboard each day next week.   The school has a plan, but is not clear how much impact the plan is having on off campus students.  The rate of infection may be at tipping point. VT claims there is no uncontrolled community spread, but the data does not support that assertion. Virginia Tech tested only on-campus students.  

VCU

VCU is showing signs of community spread.   VCU's problem, though is the space for isolation is and quarantine is near capacity.  Being in an urban setting the zip-code and regional surveillance that works well at some of the other schools is not particularly helpful here.  So, I have to rely on the dashboard.  

The With the students return, there has been a marked increase in cases.  It appears there is community spread. And it is growing.  Between the increase in the percent positive and the absolute numbers increasing, UVA is trending in the wrong direction.  UVA tested all students prior to arrival on campus.

 Watch List: empty

Other schools:

William & Mary -- so far so good -- No evidence of community spread (no cases with several thousand students on campus now.   Currently, the upperclassmen are returning to campus.   It is worth noting that pre-arrival testing has identified 20 cases, which were not allowed on campus. W&M will retest all students next week. W&M tested all students prior to arrival in town. 

CNU -- Just reopening; looks ok but the student have not been present long enough to see community spread.

ODU -- Nothing noteworth.  My concern with ODU is they did not test the students, so there may be asyptomatic/presymtomatic spreaders on campus (see JMU)

UMW -- Not open yet.

GMU -- Nothing noteworthy.  GMU tested all students.

Communities:

Based on the above charts, the issue is no longer the communities.  The issue is the campuses.  We are seeing the growth in New River (VT/Radford) and Harrisonburg.  The sad thing is those communities had the lowest numbers prior to the students return.    




Commentary:

Masks are our friends.  Masks significantly reduce the risk of disease transmission, if they are worn correctly.  The purpose of masks are 1) to contain any virus you may discharge through breathing, talking, coughing or sneezing, and 2) to protect you from incoming viruses with your breathing.  In order for the mask to function, it must cover the mouth and nostrils.  The chin and neck do not need protection (unless you have a tracheotomy).  I am seeing too many people with the nose and/or mouth exposed.  And I have even seen town employees with the masks that way.  Confronting does not seem to work, except sometimes with a neck mask, people just forgot to put it on.   It is not enough to have a mask; it most be worn properly.  WHO put the following graphic:




I try to keep my political biases out of it.  Unfortunately, some politicians have turned data into a political issue.  Yesterday, while driving home, I heard a prominent politician state that the America's numbers would be much better if we excluded NY.  I am not sure when the quote was captured.  But, I am a data and science person; based on that, I wonder how policy can follow when data is fabricated.  The data for cases in the USA, with and without NY are shown below. In April, our cases there about 50% NY, since then NY has been doing better.  Since June 1, NY has accounted for 1.5% of the cases nationally, even though it has 5.8% of the population.  So, currently, NY is outperforming the country.  Even including the early days, NY is at 7.2% the cumulative cases, about 25% higher than you would expect.  By comparison, Florida accounts for 10.2% of the cases but 6.3% of the population.  And FL should have known better.  (VA is at 2% of all cases, and 2.5% of the population).

The numbers do not lie.




Attribution:

1) You can repost / share in the entirety by forwarding the link, 2) If you want share partial content, you must receive my permission – I need to make sure you understand what I am saying. If anyone sees this work being used without attribution, please let me know as soon as possible. I am willing to have an informed discussion / debate on my approach, but I want to make sure the proper context is captured.

Links: 

Source data is from the Virginia Dept of Health COVID Site

Why I did this:  About the blog

Other Sites:  John's Hopkins

Kids can pass covid to parents: Pediatric SARS-CoV-2: Clinical Presentation, Infectivity, and Immune Responses



A fun video showing masks work, guy style:








Comments

Popular posts from this blog

Daily Status, August 29

Virginia COVID-19 Status for Sept 10

Daily Status, Sept. 11