David Salzberg's Analysis of the COVID numbers for Virginia, focusing on Northern Virginia.
Subscribe to this blog
Follow by Email
Daily Status, August 30
Get link
Facebook
Twitter
Pinterest
Email
Other Apps
Daily Status, August 30:
This morning, I decided to write a bit about my approach to analyzing the data. If you are interested, here is the link: Data Approaches
Summary/Situational Awareness
Again, today's concern are with the universities. Only one school is doing well (W & M), with no increase since the students arrived. UVA is showing increases, but probably consistent with the increased populations -- they have moved to monitor closely. Meanwhile, there is community spread is present at Virginia Tech, JMU, Radford, and VCU. More on this in the section title "Colleges". The discussion of colleges will take up most of the new discussion
Across the commonwealth, there were about 938 reported new cases . This with is slightly below the average number of cases in the last three week, but within 1 standard deviation. Regionally, all parts of VA are increasing except eastern VA. SW VA is leading the pack, mostly the result of the outbreak at Radford. It is worth noting that the areas seen growth have a much lower prevalence of the virus, and and much better testing statistics. The testing numbers across the state remain ok, but not great.
Previously, we saw a strong North-south divide, with locals south of I64 trending worse than those north. That trend continues, but is less dramatic, rather there are a few pockets in the state where things are trending bad, much of which are the smaller cities with colleges, such as Radford, Harrisonburg, etc. Outside of those isolated hotspots, the state is doing well. In fact, 20% of the cases are affiliated with a combination of Radford, JMU, VT, UVA, and VCU; those schools account for 1.5% of the population.
Regions:
The statewide numbers are no longer decreasing; they are increasing by about 1% per day, being driven by colleges mostly. We need to be cautious throughout the state. Outside of college environments, we are doing ok. The best way to describe the situation is: Yesterday, the sky was not falling and today, we are not doing great. But, as locations go, much of Virginia is improving, no overall region is particularly concerning. We are around the median for all states.
What is most noticeable is the worst part of the state per capita is SW VA.
The three week growth rates are (in fraction per day)
NOVA: 1.009
Central VA: 1.007
Hampton Roads/Eastern VA: 0.984
SW VA: 1.019
NW VA: 1.010
The state as a whole is flat.
The following charts are for the 5 regions/trends. Looking at the trends, there may be a need to create a new event for SW and NW Virginia. This is mostly due to the of case surges at Radford & JMU. Note that the total number of new cases in East/Hampton Roads is now about the same as Northern VA (though NoVA has a significantly more people).
I added a new trend line for SW virginia to reflect the return of college students. I may need to do the same for NW soon.
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.
Unfortunately, most jurisdictions are showing statistically significant increase
Fairfax Co.: 1.012
Arlington Co.: 1.019
City of Alexandria: 0.987
Prince William Co.: 1.008
Loudoun Co.: 1.005
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
If we look at the current status by zip code, we see that most areas are doing ok -- The previous hot spots are doing better, and the "blue regions" are not green The one exception is the area around Warrenton, but that is caused by a single outbreak at a long-term care facility. Also, in general, for Fairfax and Arlington County, North of US50/I66 are doing marginally better than south of I66/US50. No area with significant population is doing really well (<0.1%), but all are less than 1%.
In Fairfax County, the different communities had been following similar overall trends, with higher numbers in May, much lower numbers in late June-Early July, then slow growth to the present, except the absolute rate per capita differs as much as 3x (this is a log plot). In the last week, most of Fairfax County has continued the slow growth, with the exception of Vienna and McLean decreasing about 20% week over week. My suspicious in the variation is the result of household income: the wealthier parts of the county can afford to be more careful.
Arlington and Alexandria seem to have leveled off, which is positive.
Age Distribution:
The age distribution of cases shows growth (RT>1) among all age groups, but the working age people (young adults & middle age) are higher than younger and older groups. And once again, the Teens and Tweens are showing the slowest growth rate (practically flat). Once again, the teen bubble's social isolation from anyone not a teen is working for them. It proves if one only associate with non-infected individuals, then the risk is low. However, based on observed behaviors among that age group, if the disease expands at all, then there will be a problem.
One interesting aspect of the case numbers is a comparison of the rate of cases for teens (about 0.02% of all teens in the commonwealth) are reporting to be positive, but the numbers for returning students at colleges are more like 0.25%, suggesting that most of the teens are not being diagnosed, and probably means that only 10% of that age group are sick enough to justify testing.
Colleges:
Before I get into details, I can not find evidence of any hospitalized college student at this time in Virginia. The concern with college students are long-term impacts (unknown), and spreading to more vulnerable members of the community (like me).
Note: I have added a top level table. This is color coded to the current level of tracking. Also, in all schools, the dashboard does not seem to be updated on weekends.
School
% Pos
VDH # Case
Est # Case*
Dashboard
Case
% sick
Va Tech
1.2
69
69
21
0.2
UVA
6.8
70
72
40
0.3
JMU
17.9
155
553
190
2.6
CNU
6.1
0
0
3
0.0
UMW
9.0
0
0
0.0
Radford
62.8
278
3481
98
44.0
VCU
9.1
63
114
108
0.4
William and Mary
1.9
0
0
1
0.0
*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.
First of all, Radford is now at 278 reported cases in the last week or so. And, they have only tested about 500 people in that period (based on the ZIPCode 24141). So, they are running a 63% positive rate. That means that they are unable to test all of the people that are sick. My estimate is they are probably undercounting by a factor of 10x1. If so, that means at least 44% of the student body have been infected, though at this point, herd immunity should kick in soon. It seems that Radford has testing capacity of about 60 per day; This is the scenario that I have been worried about. Fortunately, at this time, none of the (likely) students have been hospitalized. It has been very difficult to get information beyond the state numbers, but anecdotal reports are that kids who are sick are being told not to test, just assume they have the disease (this is from comments on facebook). Note the dashboard has not been updated in several days, and is using data nearly a week old. There is very little information, and Radford has not been open with the situation. To give the university the benefit of the doubt, it is possible they have been overwhelmed. They have limited the size of gatherings. The students, though, are noticing, with classes being half empty.
JMU is open about what is going on, and had a plan to contain the virus. Unfortunately, in the last two days, there has been 155 newly identified cases in Harrisonburg. Also of concern is the percent positive: it is 17.6% in the community as a whole, and 27% on campus. Unfortunately, with the community spread, JMU is at a tipping point: they may be able to stop the spread, but it is possible it is too late.
With the return of the students, the population of Blacksburg increases by 2-3x. As such we would expect a 2-3x increase in case load. The initial surge was from testing all on-campus students prior to arrival. That was expected and is now complete. After that, it dropped back 20 about 2x the pre-student value, which is not surprising with the increase in people. The concern -- why it is yellow, is the sudden increase, which may be the first step in community spread. It is being watched closely. It is not clear if the campus new cases were the result of surveillance or symptoms. The school has a plan, and seems to be following up with it. The rate of infection so far is manageable, if they can limit the addition spread.
VCU is showing signs of community spread. But, so far, it appears to be isolated to athletes. 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 students are just returning to UVA. With the students return, there has been a marked increase in cases. It is not clear if the increase is from community spread or from the disease coming in with the students. If it is the latter, UVA could drop off the watch list.
Other schools:
William and Mary -- so far so good -- No evidence of community spread(1 case in the last week with 4000 students on campus now). Williamsburg did report 5 cases yesterday, which is above there average, but they had been below average for the week prior to today.
CNU -- Just reopening; looks ok but the student have not been present long enough to see community spread.
UMW -- Not open yet.
Communities:
In terms of the local communities around the colleges, I think we can feel confident sending our kids to college -- except for Radford. The New River Health District, which includes Radford and VT is showing extreme case growth. So far, the numbers for incoming students at Tech have been low – so, most of the growth has been from Radford. The acceleration in cases in the health district is the result of Radford's outbreak.
This means that the reported cases are showing up on campus rather than elsewhere. With 11,000 tests, they have identified about 30 infected. The concern here, though, is the infected students could cause the virus to spread to the regional community, which is not taking proper precautions. In terms of the surrounding communities, ODU is no longer a problem. W & M/ CNU are doing the best. W & M is welcomed freshmen last week, as are some of the other schools. The regional trends, which is good for W & M, CNU and ODU. VCU seems to be much of the problem around VCU.
Commentary:
I want to talk a little bit about the difference between political decisions and science. Ideally, politics is balancing the multiple facets of the problem to come up with an acceptable solution. For example, balancing the risk of the pandemic to the cost of a shut down. Science can advise, but I -- as a scientist -- am focused on what I know, which is the data and the models. The actions based on that data and model analysis are political decisions.
I am writing this before I have seen Sunday morning's data (actually, writing this Saturday Afternoon). However, my expectations are the outbreaks at Radford and JMU will appear to be worse (it is possible that Radford has peaked but because of the delays in reporting, it is not clear yet). JMU could still stop the outbreak. What they do in response to the outbreak is a different decision. The political decision will need to look at information I do not have: the capacity of the health system, the impact to the broader community, etc. There will be negative consequences on society if they just shut down: sending the sick and contagious home will spread the virus to other locations. The reality is, of the total number of students at JMU and Radford, the total death toll from the virus will probably be in the low single digits or so -- worst case if everyone gets sick. And maybe 20-200 will be hospitalized -- and this could be high.
The problem is, if you send the kids home, they will bring the virus with them. The parents are in a more vulnerable population. 50 something have a 1-2 % mortality on average, and potentially much higher with comorbidities. For example, if my offspring were to bring the virus home, I would have to isolate myself -- I have several comorbitities.
Furthermore, at the universities, many of the staff -- either professors or support staff will similarly havecomorbitities. That is the political balancing act.
The same decision the public schools had to make. The kids are not the problem. The parents, teachers, custodians, and society as a whole are what factor into the decision.
It is not just the science of one field. It is the science of the disease balanced by the social science/implications balanced by the economics.
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.
Daily Status, August 29: Today's report was late because the data are late. Typically, I report 1-2 hours after the data are published. Note: Today, I focused on the colleges. Much of the text in other sections could be reused. The figures are 100% from today, though. Summary/Situational Awareness Like yesterday, today's concern are with the universities. Only two (W & M aare showing no increase since the students arrived. Va Tech and UVA are showing increases, but probably consistent with the increased populations -- they have moved to monitor closely. Meanwhile, there is large scale community spread at JMU, Radford and possibly VCU. More on this in the section title "Colleges". The discussion of colleges will take up most of the new discussion Across the commonwealth, there were about 1217 reported new cases (including Radford). This with is slightly above the average number of cases in the last three week, but within 1 standard deviation. Regionall
Daily Status, Nov 8 Reminder: Any sections that are unchanged since yesterday are grayed out. Please be kind. I have been putting out a daily update since March. As you may have noticed, I had not put one in 10 days. I am trying to get this one out, but my energy level is low. I spent half of the last 10 days in the hospital, and had multiple procedures to fix the acute issues (bile duct blockage) so I can address the longer term issues of fighting off the cancer which is growing inside of my body. I had started working with an editor to improve the quality of the writing, but I am going to forgo that, as the process takes time, and I can not be assured that I will have the energy to complete the writing at that time. I am not sure how much longer I can keep this up. I remain concerned that the media does not understand the math well enough to accurately report the situation. However, I am confident that, on Jan 20 at 12:00 noon, the excellent and dedicated government
Daily Status, October 21 Reminder: any sections that are unchanged since yesterday are grayed out. I try to update this every day between 11:00 and 12:00. However, my updates are based on when the Virginia Department of Health updates; it takes some time in analyze and collate the data. In addition, because I my own health issues, there will be days when the update is late or does not happen. The one thing to remember is the day-to-day changes are minor. Any trend takes several days to identify, and I will be looking at the data daily (that is easy), but may not update the blog. One thing that is immensely personal to me is I need to avoid COVID-19. Like, I do not know my mortality risk exactly, but last week, I thought my Infection mortality rate was 5-10%. With active cancer, that doubles, to 10-20%. So if you see me in town, wear a mask. I try to avoid challenging people, but maybe it is time. If I can see your nostrils, you are not wearing it right. If I can see a ga
Comments
Post a Comment