Where to Begin?

It’s hard to know where to even begin today. This week the US saw it’s 100,000th COVID-19 death and yet a single death, George Floyd’s, (rightfully) consumes the nation. The two numbers, 100,000 and 1 are deeply related by the racial inequalities that exist in our nation.

Blacks in America die at a higher rate from COVID-19 than whites and already had a lower expected lifespan. Those who make under $40,000/year much more likely to have lost their jobs; Black Americans earn significantly less than their white counterparts.

Data on co-morbidities, access to healthcare, access to testing are all clear that COVID-19 has a higher impact on black communities than white ones.

In other words, the inequalities that amplify the effects of COVID-19 are the same ones that COVID-19 in turn amplifies (life span, jobs, access to healthcare).

It is impossible to imagine that a virus that amplifies the inequalities that amplify it will not make both the virus and these inequalities worse at the same time. And not just long term: this summer. Higher cases, higher deaths, higher job losses, higher co-morbidities due to lack of healthcare access due to job loss. It’s a nasty spiral.

It is unclear what impact the nationwide protests will have on the COVID outbreaks. COVID-19 research is increasingly looking at indoor settings with poor ventilation as primary locations for virus spread. At the same time, we haven’t had Americans coming together, in such close contact, for so long, day after day since mid-March. Of course, in Minneapolis, the epicenter of the riots and the protests, so many stores and businesses are closed that a different type of social distancing is now in pragmatic effect.

Numbers Going into June

For the first time in 40 days we are seeing an increase in the 7-day average of new cases per day. It is not significant on its own, but we are seeing the 7-day average of new deaths per day also begin to level off. California, Texas and Virginia seem to be driving this upward tick.

We want to be clear that this is not a major reversal from a statistical perspective. However, when the trends for new cases and new deaths aren’t headed in the correct direction and there is widespread civil unrest that is likely to make things worse we aren’t optimistic for the next few weeks.

Source: COVID Tracking Project; Chart by JM Addington
Source: COVID Tracking Project; Chart by JM Addington

Tennessee is also at a 10-day increase in new cases per day, although that number still remains low in absolute terms. It is also unclear to us what affect targeted testing has had in Tennessee that could skew this upward.

Source: COVID Tracking Project; Chart by JM Addington

The 10-county area continues to stay between 10 and 20 new cases per day. That’s mostly great.

Source: Tennessee Department of Health; Chart by JM Addington

Less great is Knox County. After dipping significantly right before Memorial Day there has been a steady rise since then. Like Tennessee, our absolute number of cases remains low, however, the convergence of less social distancing and longer test times (see below) become a trio of data points that suggest a likely rise in cases.

Source: Tennessee Department of Health; Chart by JM Addington

In addition, test result turn-around times have increased from two and a half days (4/26/2020) to nearly four days (5/17/2020). Fast tests are absolutely critical to contact tracing which in turn is critical to keeping a “slow burn” rate from turning into a hockey-stick like graph.

We don’t have great data on what is driving this in Knoxville, but we’ll move into prognostication. We’ve been mildly dismissive of “opening up,” because social distancing data plateaued well before the “lockdown,” ending. Unacast’s data shows Knox County, in particular, picking up movement by May 5th and continuing through May 26th, the last date data is available for. Those three weeks represent 1.5 “full” incubation periods or about 4 median incubation periods. In other words, the cases appear to pick up after Knoxvillians interact more, which is exactly what is expected.

Source: Unacast

Other News

Three separate unpublished studies are showing that somewhere between 1% and 20% of infected people are responsible for 80% of the overall spread. In other words, many people may barely spread COVID-19 when sick at all, a handful spread it widely. This is not new to Sars-Cov-2. Explainer video.

The CDC is warning that serological tests (antibody tests) are not accurate for any given person. This is important for Knoxville where we’ve seen advertisements for individual antibody tests. These tests are accurate a population level, not for a specific person. For instance, if you wanted to know the overall percentage of Knoxvillians infected so far a random sample of +/- 1,000 people will give you an overall percentage. But for any one person:

…less than half of those testing positive will truly have antibodies.

CDC

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, and the TN Department of Health. Most sets are updated daily

Let’s take a few minutes today to talk about cloth masks. This article is not a red/blue black/white article.

The central conclusion right up front: we don’t know. We don’t know if they help, we don’t know if they hurt, we hardly know a damn thing about cloth masks and SARS-Cov-2.

Alright, let’s dive in with a refresher that is really important.

Science Moves Fast; Sciences Moves Slow

From Hippocrates nearly 2,400 years ago until the late 1800’s there were nearly no advances made in clinical medicine. [1] Until the 1600’s there wasn’t even a scientific method used to figure out if our theories on how the world worked were even right!

In a nutshell, the scientific method is as follows:

  1. Define a question
  2. Gather information and resources (observe)
  3. Form an explanatory hypothesis
  4. Test the hypothesis by performing an experiment and collecting data in a reproducible manner
  5. Analyze the data
  6. Interpret the data and draw conclusions that serve as a starting point for new hypothesis
  7. Publish results
  8. Retest (frequently done by other scientists)

Do not skip over #8. At the point where you can reproduce a study (steps 1-7) and get the same or similar results is that point at which your hypothesis (#3) becomes theory and part of accepted scientific knowledge. A single published study does not a scientific fact make.

We live in a world where due to increased communication, resources and technology we are able to make advances in science faster than ever before. At the same time, this typically happens over years (or decades) of running the questions through the scientific method. #8 rarely, if ever, happens in weeks. Which, compared to 2,400 years of Hippocratic stagnation isn’t bad.

The Discussion

Between #7 and #8 there is really a step #7b, where scientists basically rip apart each other’s work. Often, this is a sign of respect, your study was good enough to draw attention from other top scientists. This is the point where uncertainty is identified (not resolved) and typically more questions are thrown into the mix to be answered when the next person goes to try to reproduce the study. This discussion, while transparent in scientific journals, is typically done without the public realizing it happened at all.

This is normal. However, at present, this discussion is playing out with much more public visibility than we are used to.

The Tests & Data (#4 & #5)

Outside of medicine and other sociological fields this is pretty straightforward. You need to gather data and test a hypothesis on mice? Not hard to come by mice or do your test. Medicine constantly raises thorny ethical issues and sometimes you just don’t have data available even if the ethics are laid aside. Sometimes, you need to wait for an opportunity to appear to test appropriately.

Cloth Masks & SARS-Cov-2

Which brings us to cloth masks. You know what we have in western countries? We have N95 respirators and surgical masks in great supply…. up until January 2020. You get easily get an N95 mask at The Home Depot for your own home improvement project up until a couple months ago. They were cheap and plentiful.

Given that high quality masks were cheap and in great supply (and well tested) you know what we didn’t look at? We didn’t even think to the 100 different ways laypeople could create, wear and implement cloth masks. Just wasn’t an issue.

The first paper we at JM Addington even saw on it was titled, “Rapid Expert Consultation on the Effectiveness of Fabric Masks for the COVID-19 Pandemic.” Rapid means, “we did this as fast as we could,” “Expert consultation means,” we didn’t actually do a study — we looked at some other stuff.

This paper had, in our view, two primarily conclusions. Conclusion #1:

Therefore, we have only limited, indirect evidence regarding the effectiveness of such masks for protecting others, when made and worn by the general public on a regular basis. That evidence comes primarily from laboratory studies testing the effectiveness of different materials at capturing particles of different sizes.

I.e., we don’t know much, and what we do know is not from real world scenarios.

Conclusion #2:

The evidence from these laboratory filtration studies suggests that such fabric masks may reduce the transmission of larger respiratory droplets.

In short, “cloth masks maybe help based on looking at some fast non-real world studies but we really aren’t sure.

And we’re not even diving into into the issues of:

  • Small samples sizes
  • Wearing masks right/wrong
  • Whether blocking large droplets is effective in reducing transmissions [2]
  • No reproduction of results (step #8)

Are Cloth Masks Bunk?

A well-trained scientist would probably say that wearing a cloth mask or not wearing one is not even wrong yet. To say, “we don’t have evidence,” means that we don’t have well-designed controlled studies that have been re-produced to show that we know we are right. Scientists don’t like to say that they know until they know that they know that they know.

So: cloth masks: the jury is still out.

What about Particle Sizes?

We’ve seen a number of posts proclaiming that cloth masks can’t work based on the size of the gaps in the mask versus the size of Sars-Cov-2 particles. First, this argument completely bypasses the scientific method where we test ideas. Second, it still isn’t sufficient to render masks useless. To steal an analogy, imagine an NFL team trying to run onto the field through a bedroom-sized door: while they can each fit through they can’t all run out together at the same speed. Masks don’t need to be 100% effective in blocking virus particles to still prevent enough from getting through to reach the level needed to infect another person.

Why Are We Being Asked to Wear Cloth Masks?

Cloth masks are wholesale thinking, not retail. The goal isn’t to protect any one person: it is to drive down the effective reproductive number, aka Rt. Rt measures what the current transmission rate of the virus is across a population. (It is not the same thing as R0.) In the early stages of an epidemic, the goal is typically to contain the epidemic before it gets bigger and becomes a pandemic. For examples, SARS, MERS and Ebola have all been effectively contained epidemics.

At this point, public health strategy has turned to mitigation, trying to slow the virus’ growth. Stopping it outright has become effectively impossible.

No one mitigation method will be or needs to be a silver bullet, the strategy is that if enough of the population practices enough of the measure you can drive the Rt number low enough that the epidemic locally either stays stable or actually decreases. Mask wearing is one of these measures. It doesn’t need to be anywhere near 100% effective by itself in order to drive the Rt number low enough in combination with other measures to make a real difference.

Should I Wear a Cloth Mask?

If we got into the amount of uncertainty that exists around social distancing, what the Sars-COV-2 is going to do on its on you’d feel really crazy by the end. We don’t pretend to have an answer.

Your author will give his personal view. It doesn’t hurt me to wear a mask, and I believe that I have one of the better ones that could reduce transmission at any rate. The purpose of all the things we are doing, social distancing, masks, limited retail/restaurant/church capacity is meant to slow SARS-Cov-2 so that the overall pandemic stays flat or goes down. No one thing other than God himself eliminates it.

If wearing a mask inside Target, Kroger, my church, at my clients’ places of businesses, helps to keep the virus manageable so that we can continue to live normal-ish lives I’m all for that. I own a business that makes more money in a normal economy, I have eight kids, I don’t like being sick, I like eating out at restaurants. For all these reasons if keeping the virus in check is important to me. Physically and financially.

If you look at the data and think it doesn’t make enough of a difference to wear one I also don’t begrudge you. You are right, the jury is still out.

At the end of the day, it is our collective responsibility to remember that it is humanity against the virus: not dems vs reps, red vs blue, or mask vs no-mask. It is up to all of us to do our part to fight this damned thing from hell however we see fit and condemn the virus and love our brothers and sisters.

[1] The Great Influenza: The Story of the Deadliest Pandemic in History

[2] I haven’t seen anyone who thinks a cloth mask significantly reduces aerosol transmissions, but it isn’t clear what role these smaller droplets play in transmission.

Short post, economics and the CDC is being less precise with their numbers than the media.

Thursday Economic Update

New initial unemployment claims were released this morning (like every Thursday), they “fell” to 2.4 million, or about 10x what you’d expect to see in an average week in January or February of this year. The official unemployment rate still is around 15% but this is surely above that by now. Fortune estimates it to be 22.5%.

Source: FRED

If you were studying economics 20 years ago you would have been taught that 5% was the lowest the economists typically thought unemployment could be. In recent years that’s nudged down, maybe to 4%. So 15% is really, really high, as is 2.4 million initial jobless claims.

Our take is that even with limited re-opening it shouldn’t be unexpected that jobless claims lag the pandemic:

  • The PPP program (free money to small business) surely put off some, but not all claims
  • A number of businesses would opt to keep people on payroll as long as is tenable
  • Other businesses are just getting to the point where fixed costs outstrip their ability to operate at regular employment

If new claims are still this high in a month that will be a significant problem.

Fed chairman Jerome Powell was on 60 Minutes last weekend and Axios gave us some of the highlights:

Fed Chair Jay Powell, in a very interesting “60 Minutes” interview with Scott Pelley, made several points that should be on your radar (videotranscript):

On what metrics he watches hour-by-hour: “[A]t the moment, the thing that matters more than anything else is the medical metrics, frankly.”

On whether the Fed sees unemployment of 20% to 25%: “[T]hose numbers sound about right for what the peak may be.”

On whether he thinks history will call this a Second Great Depression: “I don’t think that’s a likely outcome at all. There’re some very fundamental differences. … [W]e had a very healthy economy two months ago. And this is an outside event — it is a natural disaster, in effect.”

We’ll note that we’ve projected a 25-30% peak and posted several times on how different this is from any other economic event the modern economy has ever gone through.

For what it’s worth, Bank of America has said that they think the initial downturn was worse than they initially thought, but are forecasting a faster recovery. The thing to keep in mind about any of this big picture economic forecasts is that the economics downturn will hit different people, industries, and geographies very differently. Those making under $40,000 are more likely to have lost jobs than those making over $80,000, Walmart, Target & Amazon are powering through this pandemic, Uber and Delta are not, nor is JC Penny. he stock market seems to have lost nearly all mooring to the “fundamentals.”

So, these forecasts are helpful to know how things play out on a national, and almost conceptual level, but your specific economic experience is not tethered to the national economy.

Oregon’s stay-at-home order was thrown out, and then the toss out was “stayed,” both on Monday. You can expect to see more of these court challenges.

The biggest COVID-19 headline we’ve seen today is this report from The Atlantic that the CDC and some states are combining the results of their PCR and serological tests in reporting, including two of the states that we reported on opening up early, Texas and Georgia. These are two very different tests. It would be like combining the scores of your child’s reading test with that of his or her class and then just reporting that final score. Or counting your apples and oranges and just reporting how many total pieces of fruit you counted.

This really makes the testing numbers coming from CDC, Texas, Georgia, Pennsylvania and Vermont close to worthless.

To be clear, doing both tests are good. Reporting both results is great. Reporting those results together is bad.

One thing you’ll hear me say around our Microsoft Teams “office” a lot is that “everything is backwards in COVID world.” This is another example of that, where CDC — the gold standard of the entire world for public health — has less reliable numbers in the middle of a pandemic than John Hopkins or The COVID Tracking Project, which is led by a magazine (The Atlantic).

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily

Today, after a few days of agitation from local lawmakers, Knox County changed their re-opening plans. We’ll discuss that and then hit numbers.

Knox County Re-Opening

Knox County announced today that Phase 2 will come three days early. This comes five days after the state of Tennessee announced that restaurants and retail could return to full capacity May 22nd, and after repeated statements from local lawmakers pushing KCHD to move to Phase 2 early.

There are there several ways to look at this, and all of them are important. First, opening 3 days early with essentially flat numbers isn’t likely to make a difference in COVID-19 spread that we’ll ever be able to measure. Second, this push to match the state’s re-opening guidelines cedes influence on the topic to the state. It’s an odd viewpoint coming from Mayor Jacobs, in particular, to cede power up the chain instead of keeping decision making local.

However, the lack of a unified plan across the state also made this inevitable to a degree: any time an industry in one county has an advantage over an industry in another, businesses are fearful of missing out. The economics of the situation pressure everyone to open up as fast as the fastest county/state.

Third, given that Tennessee is already re-opening restaurant and retail capacity to 100% this week and that the rest of the guidelines are, essentially, unenforceable, the final cessation of power and influence effectively is to individuals and businesses. The government will give us suggestions: it is up to us whether or not to follow them.

This is a realpolitik — short of turning into a police state the governmental shutdowns rely primarily on voluntarily compliance. In a state that hasn’t been hit hard by COVID-19 cases and deaths it is clear that patience for that compliance has waned considerably. It also doesn’t mean that the state, or the county, is back to a pre-pandemic normal.

The public orders matter far less than what people choose to do, on their own, voluntarily. It also remains up to individual persons to wear masks, wash hands, physically distance while out, etc. It also remains up to them to choose to dine-in, or do curbside pickup, at their favorite restaurant. Put another way, opening up is not a yes/no binary choice nor were the lockdown. Or another, the virus cares about what people do not what regulations politicians put in place.

Finally, there is a humility that we don’t know in full what causes pandemics to ebb and flow. This post goes over three very different scenarios painted by some of the most prominent epidemiologists in the country. It is possible that COVID-19 was already going to recede on its own and that lockdowns/opening-up play less important of a role than we realize. We may not know for weeks or months.

The Americans are a curious, original people. They know how to govern themselves, but nobody else can govern them.

A British officer leaving New York, 1783

We have earned this in part, now let us earn it in full.

How is JM Addington Reacting?

We haven’t changed plans yet. Our views for our business are:

  • A conservative approach (work from home, wearng masks, etc.) is not likely to lose any business and we will probably pick some up from people that appreciate it.
  • The same is true for employees: everyone feels appropriately safe in our current operations.
  • While the risk of several of us getting sick at once is low, it would be devastating to our company. Our best “insurance” against it is to stay pragmatically cautious.
  • We believe that there are 2-4 good months ahead of us, as business picks up its pace and some businesses want to invest the liquidity injections they just got from PPP and EIDL. We expect opportunities over this summer that we haven’t seen in years.
  • We’re trying to keep a maximum range of options open for the fall and winter. History strongly suggests a second wave will come. If that happens we want to have done our best in the good months to be prepared to ride it out without counting on another government bailout.

Overall, it’s remarkably unremarkable. We are, essentially holding stead since mid-April in community cases. Our testing remains high enough that we are testing roughly twice the target. And even as we target populations where we expect to see more cases (prisons, nursing homes) the testing ratio remains great.

Local News

Wampler’s Farm Sausage, which reported a single asymptotic case a couple weeks ago only found one more after extensive testing (we think, their entire workforce.) This speaks very highly of Tennessee’s testing, that we are now able to avoid such an outbreak.

…but…

Monterey Mushrooms, located in the same city, had an outbreak of at least 59 employees of about 300 over the weekend. We have no way to verify this, but we would be surprised based on other outbreaks if there was not a housing aspect that played a role here as well. (I.e., multiple workers living in close quarters in the same housing areas.)

Numbers

Nationally, numbers are still stable on average. (The uptick at the end is a result of a particularly low day falling off the average.)

US COVID-19 Cases and 7 Day Rolling Average by JM Addington Technology Solutions in Knoxville, TN
Source: Data from The COVID Tracking Project; Chart by JM Addington

However, that blended average hides more nuanced sub-plots to the story. Headed up in cases include AR, AZ, ME, MI, MN, NC (sharply!), ND, OK and TX. If we change the chart to just those states you don’t see that there ever was a peak. Don’t take this beyond its literal meaning: each region/state needs to be considered on it’s own, not just as part of the US average.

Selected States COVID-19 Cases by JM Addington Technology Solutions in Knoxville TN
Source: Data from The COVID Tracking Project; Chart by JM Addington

Tennessee remains mostly flat, up a little bit this month, we believe, from targeted tests in prisons and nursing homes.

TN COVID-19 Cases and 7 Day Rolling Average by JM Addington Technology Solutions in Knoxville, TN
Source: Data from The COVID Tracking Project; Chart by JM Addington

The county area also remains stable, except for the major spike from Monterey Mushrooms. The real question here is if this will fuel a larger outbreak of if it was tested in time to be contained. 3-4 weeks will tell.

Area COVID-19 Cases and 7 Day Rolling Average by JM Addington Technology Solutions in Knoxville, TN
Source: TN Department of Health; Chart by JM Addington

Finally, Knox County itself also remains very stable. More importantly, our hospital system is still within capacity.

By the way, if you want to see the same graphs for your state click here, we pull nearly all of our own visuals from this report. Typically updated at the same time the evening’s post goes up.

Here are GA new cases from about a month ago.

Also, you can do it on your phone but it is much easier to navigate on a bigger screen.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily

I get these emails with some regularity. I’ve kept my same email address for about ten years now and so there are a lot of chances for my records to be exposed.

In this latest breach, over 22 million records were exposed. And unlike most breaches, it was more than emails and addresses. This included mobile phone numbers, CRM entries, records of real world interactions, summaries of legal briefs, and more.

In fact, at this point my email address and password have been exposed multiple times online. Yours have, too, you just don’t know it yet.

This is a huge part of why we agitate for:

  • Using multifactor authentication
  • Disallowing password re-use

Even though my data is out there, yet again, the combination of using multi-factor authentication and different passwords for different sites nearly guarantees that I won’t suffer a personal breach. In fact, both Microsoft and Google have reported that turning on multifactor authentication stops 99% of all breach attempts.

Next time we tell you that multifactor authentication needs to be turned on, you know why!

Today we’ll look at three Southern states that opened around the same time and compare how they are doing.

Re-Opening

Tennessee is now 17 days into re-opening, as is Texas. Georgia got the jump on us and began their re-opening on April 24th, a week earlier. Of those three states, Texas probably had the most conservative plan, Tennessee right behind it, while Georgia’s plan was… ambitious.

So, this far in what does the data say?

We’re going to start with one giant caveat, in many ways a post like this ought to way until a full three or four week have passed to make judgements. However, our belief is that in each of these three states the data is consistent enough in its trends to start making calls, even if its early.

[Edit: 5/18/20] A thoughtful reader (thanks Rob!) points out the charts below could be misread. To be clear, even though we have highlighted the entire opening period, only the most few recent days would have cases that began during the re-opening period.

Tennessee Re-Opens

Tennessee began a trend up at the beginning of the re-opening period, on that cannot be attributed to re-opening and is most likely from targeted testing that began at correctional facilities. Last week’s bump is attributable to the same thing.

Overall, it’s remarkably unremarkable. We are, essentially holding stead since mid-April in community cases. Our testing remains high enough that we are testing roughly twice the target. And even as we target populations where we expect to see more cases (prisons, nursing homes) the testing ratio remains great.

Source: Data from the COVID Tracking Project; Chart by JM Addington

The cases that we do have remain around Nashville and Memphis. This map, which you won’t find elsewhere, shows cases for just the last 2 weeks by county in Tennessee. The one part that isn’t clear is that Davidson and Shelbyville are much higher than the surrounding counties, even though they come out in similar colors.

Source: Data from TN Department of Health; Map by JM Addington

This map matters because if you look at total cases, ever, then much more of the state looks awful.

For the local situation we are going to introduce another new chart, new cases by county by day. And in short, the Knox County region is just steady. We peaked in mid-April but just never had an awful spike. In fact, in the last 7 days Knox County has only seen a new of 28 new cases.

Source: Data from TN Department of Health; Chart by JM Addington

To prognosticate: we don’t believe that anywhere in the US can or will see a caseload lower than Knox County’s for any sustained period while COVID-19 is still circulating widely. We really think that this is as good as it gets.

So, Tennessee seems to be doing really well on re-opening.

Georgia Re-Opens

Until Wisconsin threw out their entire stay-at-home order and filled up the bars in a single night Georgia really held the title of most ambitious re-opening plan. Candidly, even “plan” seems like too kind of a word to use. So, how has Georgia done, with a one week head start?

Quite well. Like Tennessee, Georgia shows stability more than anything. not a sharp decrease but no explosive outbreak here either. Their testing continues to ramp up and Georgia also does about twice as many tests as is targeted. Note that their new cases per day is roughly double Tennessee’s.

Source: Data from the COVID Tracking Project; Chart by JM Addington

Their primary outbreak is constrained to a single geography: around Atlanta, with 36% of all cases:

Source: Georgia Department of Public Health

Those five counties are also showing a decrease in cases over the last two weeks. Incredibly, that decrease becomes much more pronounced the longer Georgia has been opened up.

Source: Georgia Department of Public Health

Georgia has been critiqued for having data lags that last up to a couple of weeks, so it is possible that we’ll see the data for early May change. However, the data we have now shows that Georgia also is opening up alright.

Texas Reopens

So, the South is two for two so far, what about Texas? Their stay at home order expired the same date as Tennessee’s and their business capacity was held to 25%, half of Tennessee’s. Although Texas also allowed for some more industries to re-open.

Texas is where the successful re-opening story goes to die. They went from a downward trend in Late April to an uptick in new cases that began at least a week before opening up even took place, and that hasn’t abated since. Their new cases per day (average) are roughly triple Tennessee’s (with about 4x the population). And this is even though Texas is also testing about double the testing target.

Source: Data from the COVID Tracking Project; Chart by JM Addington

Not shown are deaths, which are ticking up in Texas since last April as well. If there is a mitigating factor is it that Texas has nearly 30 million people. Also, to keep things in perspective their peak is still over 30% lower than New York’s lowest point post-peak, and New York state has less than two thirds as many people.

It’s possible that there are specific factors driving the outbreak that we’re unaware of — we haven’t been tracking Texas at JM Addington — but regardless it’s clear that they are having a tougher time of it than Tennessee & Georgia.

What does this mean more broadly?

  1. There is not a single narrative that explains how the US is handling COVID-19. We’re declining at a nation, with some states holding steady of quickly declining while others move the opposite direction. New York’s best day is still worse than most state’s worst day.
  2. Opening up has more than one potential outcome: A lot of us are looking at Wisconsin now after their supreme court struck down the statewide stay-at-home order and Wisconsinites quickly filled the bars. However, from our brief analysis here it’s clear that doesn’t determine which was Wisconsin will swing.
  3. There is more we don’t know than what we do know: it’s hard to lay claim to victory where things are going well when it’s not clear what causes them to go so well in one place and move in the opposite direction in another.

By the way, if you want to see the same graphs for your state click here, we pull nearly all of our own visuals from this report. Typically updated at the same time the evening’s post goes up.

Here are GA new cases from about a month ago.

Also, you can do it on your phone but it is much easier to navigate on a bigger screen.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily

First, we will go over the top two COVID-19 views we believe are being left out of the public discussion, second, hit national and regional headlines and finally go over numbers.

If this page is too long for you, read until “News” and stop there.

One Theme in Two Areas

The public discourse this week is starting to look at the longer view of COVID-19 in both health terms and economic terms. The most balanced view and, we believe, most correct view is that these are intractably the same thing. The worse the health outcome, the worse the economic outcome. (If you haven’t read our post from Sunday that goes over three major COVID-19 scenarios we’d encourage you to visit that section now.)

This week we’ve seen Dr. Fauci warn that states that “skip over” opening criteria risk longer-term setbacks. Dr. Fauci has been nothing if not consistent in articulating this, here he is three weeks apart discussing states skipping steps as they open up:

This should be viewed as an economic setback as much as a health outcome. This isn’t crazy talk, the official criteria boil down to, “COVID isn’t getting worse and your state’s healthcare system can handle the cases;” if COVID is getting worse and your healthcare system can’t handle how can you actually expect the economy to produce?

Today, the chair of the federal reserve cautioned that, “the recovery may take some time to gather momentum,” which is economist speak to say that things are going to be bad for awhile. How long and how bad? That’s still hard to say, “the scope and speed of this downturn are without modern precedent…”

“Without modern precedent,” is a statement we 100% agree with: modern economics more or less date back to the founding of America and the modern economy with globalism, worldwide capitalism, interlinked economies, etc., dates back decades not centuries. So for as much as COVID-19 is the first modern health pandemic of this scope and size, so is the economic downturn.

So, to the extent that we can’t get COVID-19 under control both nationally and globally we’re going to have trouble getting the economy going.

The best COVID-19 current models have things getting better, but very, very slowly. We had a steep tick up, like a roller coaster climb, but we are missing all of the dramatic fall.

We’ve said before and we’ll say it again: we’ll all get through this, this isn’t likely to turn into the next Great Depression. But getting a handle on COVID-19 sure would put us in a better economic state.

Economic Data

We hit unemployment Sunday, and it’s bad, at nearly 15% with the real number likely closer to 25%. Quarter one GDP (overall economic activity) dropped nearly 5%, assuming that most of the drop occurred in March as stay-at-home orders first hit we’re probably looking at a 15% decline in economic activity inside a single month.

Tennessee released revenue figures for April, yesterday, May 12th. These figures largely reflect changes in March as March sales generate April income in taxes. (As a business, any sales tax we collect has to be paid by the 20th of the next month.) Sales tax, specifically, is down 6% from a year ago

Keep in mind that economic data lags real life, so it is likely that we’ll see the worst economic numbers at the same time we’re seeing the best health numbers.

News

Tennessee is going to lead our news today. First, Vanderbilt came out and updated their COVID-19 report for Tennessee. Here are the top takeaways and our brief analysis:

  • Hospitals are not above capacity: Keeping the healthcare system inside capacity was the biggest driver of lockdown, so this is critical.
  • Tennessee is either plateaued or barely declining: This is mostly good. However, the decline is so slight that we don’t have room to mess it up right now.
  • Their first report was overly pessimistic: Modeling is hard when you deeply understand the disease and have great data. No one in the world has more than 5 months of experience with COVID-19 and 6-10 weeks of data is not a lot of data for modeling. The Vanderbilt team has done a great job using their expertise, staying humble about what they know and being transparent about what has changed.

May 8th, the Tennessee Lookout reported that police and sheriff’s offices across the state have access to the names and addresses of people have who have tested positive with COVID-19. Yesterday, the City of Knoxville quickly discontinued the practice of receiving this information from the state while the Knox County Sheriff’s Office doubled down.

This is not one of the things Tennessee is getting right.

  • Law enforcement should be treating every encounter as if it is with a COVID-19 positive person
  • It doesn’t help Governor Lee’s mantra of, “if in doubt, get a test”
  • It is likely to discourage testing among some communities where testing is needed most

I still believe in having access to that list. I think it has huge benefit… But… I think if it kept people from getting tested it was not worth it.”

Eve Thomas, Knoxville Chief of POlice

Thailand, the second country to deal with COVID-19, has no new cases today. Thailand has a population of nearly 7 million and has recorded only about 3,000 COVID-19 cases, a quarter as many as South Korea and about a fifth as many as all of Tennessee.

A new study on hydroxychloroquine, from New York City, shows it to essentially be ineffective for COVID-19.

Numbers

Numbers are not materially different from Sunday, so we are just going to show two charts, national cases and Tennessee. While the US continues to show a downward trend in new cases it continues to be a very slight decline discounting New York. Adjusting for the results from mass prison testing, Tennessee is essentially flat over the last 6 weeks (unadjusted data shown).

Source: Data from The COVID Tracking Project; Chart by JM Addington
Source: Data from The COVID Tracking Project; Chart by JM Addington

What Happened to the IMHE Model?

That seems to have quickly fallen out of favor with many (most?) medical professionals. It overestimated deaths on the way up and underestimated them on the way down. We’ll show you how their second round of modeling did, but, at this point, don’t see a reason to continue to reference it. That said, IMHE has also done a world-class job in modeling transparency around their modeling, we just believe that there are better models out there right now.

The pink line should be tracking the middle blue line:

Source: Data from The COVID Tracking Project and IMHE; Chart by JM Addington

By the way, if you want to see the same graphs for your state click here, we pull nearly all of our own visuals from this report. Typically updated at the same time the evening’s post goes up.

Here are GA new cases.

Also, you can do it on your phone but it is much easier to navigate on a bigger screen.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily

First, we will go over the top two COVID-19 views we believe are being left out of the public discussion, second, hit national and regional headlines and finally go over numbers.

If this page is too long for you, read until “News” and stop there.

View 1: It is a Long Road Ahead

As things begin to “open up” it is easy to get lost in the feeling that we’re somehow over the hump or the worst of the virus. This is, at best, half true.

First, US cases and deaths, if you discount New York, are either plateaued or showing a mild decrease — and then for only about a week. (See Numbers below)

Globally, cases and deaths simply continue to rise.

Second, there is no indication that COVID-19 will slow its spread in a permanent way until we hit 60-70% of the population infected. If you assume that about 5% of the US population has been infected and consider our current death toll (~75k for our napkin math here) you end up with:

(60% / 5%) x 75,000 deaths = 900,000 deaths left to go, give or take.

(The most optimistic estimate would be about 300,000 deaths total, the most pessimistic about 1,000,000 deaths total).

Or, simply consider how many more “Aprils” we have to go through.

Global COVID-19 Total Cases and Deaths
Source: Worldometers

Three of America’s leading epidemiologists, Kristine Moore of the University of Minnesota, Marc Lipstich of Harvard, Michael Osterholm of the University of Minnesota and John Barry — a historian of the 1918 flu pandemic, co-wrote a paper that goes over the three scenarios that they think are the most likely for increased infection.

In Scenario 1, we continue to go through, “a series of repetitive smaller waves
that occur… over a 1- to 2-year period, gradually diminishing sometime in 2021.”

This, “Peaks and Valleys,” scenario doesn’t assume that all places experience COVID-19 equally. While the first “wave” hit New York, Chicago, Louisiana, etc., a second wave may hit Denver and Austin, a third Atalanta and Minneapolis, etc.

Source: COVID-19: THE CIDRAP VIEWPOINT

In the second scenario, COVID-19 actually lets up over the summer (the second dip in the graph) and “is followed by a larger wave in the fall or winter of 2020 and one or more smaller subsequent waves in 2021.”

This is what was seen in both the 1918-1919 and 1957-1958 flu pandemics. Dr. Osterholm describes this as the worst scenario, and one where we possibly have the heath care system overrun.

Source: COVID-19: THE CIDRAP VIEWPOINT

In the third scenario, our current spring peak “is followed by a “slow burn” of ongoing transmission and case occurrence, but without a clear wave pattern.”

This scenario, like the first, assumes that different places experience their own outbreaks at different times.

Source: COVID-19: THE CIDRAP VIEWPOINT

The report ends with the one thing we wish every American was aware of right now:

[W]e must be prepared for at least another 18 to 24 months of significant COVID-19 activity…

COVID-19: The CIDRAP Viewpoint

This is critically important if we want to learn how to live with the virus, opened-up. Planning on the family level, the work level, the city/county/state/country levels are all crucial to us going forward as we get through this instead of finding ourselves frozen.

View 2: Unlit Powder Still Explodes

In places such as East Tennessee, among others, it feels right now like we dodged a bullet, that things are safe now. You can see it most clearly when people don’t take advice from public health authorities. With about 1 case for every 2,000 people it is an easy view to take.

It’s also incredibly wrong.

We are in a storeroom of 50 dry gunpowder kegs where a handful of them have exploded, just not ours. However, the fire still smolders and we are still a keg of dry powder.

And that’s the state we live in until the fire is out.

News

The most important news item of the last few days goes to a new antigen test that the FDA has granted emergency approval for. It looks like it’s great for reporting positives, for negatives a PCR test is recommended to confirm. I.e., it can tell you if you have COVID-19, however, if you test negative there is still a 1/5 chance you have COVID-19. Cheap, 15 minute turn-around times. Of course, it still needs to be mass-produced.

Unemployment is up to nearly 15% and Treasury Secretary Steven Mnuchin said, “the reported numbers are probably going to get worse before they get better.” Our guestimate is that it wouldn’t be surprising to see that double in the short term (2-3 months). Remember, we asked everyone to stay home for a few weeks: these numbers aren’t yet surprising.

Vice President Pence’s press secretary and at least one other White House aid tested positive for COVID-19. It will be interesting to see how this type of low-level exposure will affect leadership at all levels of governments, and organizations. Do top leaders need to self-isolate every time they have had potential contact with a COVID-19 positive person? (Pence is not) If the leader never gets infected, that is potentially a never-ending string of self-quarantines. (True for any person, not just leaders.)

According to Harvard, Tennessee is one of 7 states doing enough testing. They use the same positive test ratio that we’ve referenced here, a 10% positivity rate. Also, they use 7-day rolling averages for tests and new cases as we have.

Note: Tennessee could easily go over 10% this month because of the targeted testing that the state is conducting. This is where context becomes incredibly important: if 15,000 Tennessee inmates test positive in a matter of 2-3 weeks but most of them are asymptomatic, is 10% still the right target number? Or do you keep the 10% target for community transmission but not prison transmission? We expect to see state leaders approaching these as separate topics from a health perspective.

Restaurants open tomorrow in Nashville.

Tennessee is testing all of their prisons after one prison tested positive at a 50% rate; 98% were asymptomatic.

The Friday KCHD briefing confirmed that elective surgeries (non-emergency surgeries) will resume in a phased-in manner, starting with outpatient surgeries. Dr. Buchanan also said that Knox County is working through several cluster events (at least 2 COVID-19 cases together) but attributed all of them to families.

An asymptomatic employee at Wampler’s Farm Sausage in Lenoir City, TN tested positive for COVID-19, Friday. We’ll see what happens here, clearly, meat processing plants have been a major source of outbreaks in other areas of the country.

Numbers

Nationally cases appear to be declining — and they are — but, as we’ve reported for weeks here, the story in New York is different than the story elsewhere.

US COVID-19 Cases and 7 Day Rolling Average, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

If you remove New York completely, we are maybe beginning a decline. There certainly is not enough of a streak to say that definitively. Sounds like what we said last week.

US without New York COVID-19 Cases and 7 Day Rolling Average, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

Deaths look pretty much the same: the curve is very, very flat, and about tied for the leading cause of death in the United States. (If you included probable deaths it would be the current leading cause of death far and away.)

US New York COVID-19 Deaths and 7 Day Rolling Average, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

Like cases, taking out New York takes out any notable decline:

US without New York COVID-19 Deaths and 7 Day Rolling Average, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

There are many different correct ways to look at this chart of deaths. Here are what we think are the most important:

  • Going back to our beginning analysis, it’s a long road ahead
  • The overall situation isn’t better than it was a month a ago, it just isn’t worse either
  • Until a vaccine comes this is what the path to herd immunity looks like: a new killer of Americans

For the first time in our reporting the positivity rate in the US has fallen under the 10% threshold.

US COVID-19 Positivity Rate, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

Moving on to Tennessee, which remains an outlier (aka unlit powder keg):

TN COVID-19 Cases and 7 Day Rolling Average, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

And as we reported above, testing is a real bright spot for how Tennessee has handled COVID-19 so far. We believe that jump is a set of results from a prison coming back, otherwise we might be lower. However, Tennessee is handily beating the 10% target currently.

TN COVID-19 Positivity Rate, May 10th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

Sub-regionally our case count is steady, a very low burn. Yes, it appears to take an impossible step back. Typically this happens when a case is incorrectly registered inthe wrong county and then subsequently “transferred” to the correct county.

May 10th, 2020 COVID-19 cases for Anderson County, Blount County, Grainger County, Jefferson County, Knox County, Loudon County, Roane County, Scott County, Sevier County and Union County
Source: Data from The New York Times; Chart by JM Addington

Knox County numbers:

May 10th, 2020 Knox County COVID-19 Cases
Source: Knox County Health Department

Knox County Health Department is also releasing much more data on their website, this one shows the lag between when a test is taken and when the result is returned. It is down to 2.6 days, the right direction but still a day and half too long:

Knox County Test Result Lag Times
Source: Knox County Health Department

Finally, we have cases by ZIP code in Knox County, a map we show because you want to see it. Take it with an entire salt shaker, results may be as much about population density and testing access as bona fide cases or disease burden.

May 10th, 2020 Knox County COVID-19 Cases by ZIP Code
Source: Knox County Health Department

Of course, even with those caveats, it’s hard to imagine that opening up won’t bring more cases, and deaths.

Knox County remains in its own little world. One of the most important aspects of our low growth in the absolute number of new cases today is that our health department has enough to staff to do testing and contact tracing. After you get above a certain point your percentage growth rate can be slow, or even negative, but you still don’t have enough people to fight it. The cat is still in the bag here.

By the way, if you want to see the same graphs for your state click here, we pull nearly all of our own visuals from this report. Typically updated at the same time the evening’s post goes up.

Here are GA new cases.

Also, you can do it on your phone but it is much easier to navigate on a bigger screen.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily

Today we’re going to go over the two main types of testing, and why we can’t just jump to serology testing.

How does increased testing affect New Cases?

Let’s start with a point we made yesterday:

There are two primary reasons that we test in public health:

  1. To find out if a specific individual is sick. In the case of COVID-19 this would be to ask them — and anybody that they have been in contact with — to self-isolate to prevent further spread
  2. To find out how widespread a disease is in a community or geography (city, county, state, prison, hospital, nursing home, etc.)

There are two primary types of tests for COVID-19, at this point, essentially, one for each.

Test Types

This first time of testing makes up all of the confirmed cases that we know about in the US and is done exclusively (right now, as far as we know) by RT-PCR. This is the test where they put a swab far enough up your nose to tickle your brain, then use chemicals (reagents) to transfer the sample to a test tube or similar. After that, chemicals are added to convert virus’ RNA to DNA. Finally, the DNA is “amplified,” — increased — until there is enough of it to identify.

The RT-PCR test typically has a high degree of accuracy of identifying a sample where the virus really does exist, and successfully not identifying samples where the virus does not exist.

The RT-PCR test mostly tells you if you are sick right now. The RT-PCR testing is limited right now because the world doesn’t have enough of all of the supplies required — swabs, PPE, reagents, etc. — to test as many people as we’d like.

The RT-PCR test is also critical to contact tracing: finding everyone that has been exposed to someone currently infectious. If I expose you and the health department tests me positive and then calls you right away to let you know that you have been exposed you can stay home and stop the spread. Other tests — to date — don’t offer this kind of immediate payoff.

The type of test you may have heard about in the news — the antibody test or serology test — is a blood test that looks for antibodies, proteins that your immune system creates to fight off invading viruses and bacteria. The antibodies stick around after an infection, sometimes for a little while sometimes for the rest of your life.

The serology tests that we’ve seen so far are typically in the 95-98% accuracy range. To use an example from Dr. Osterholm, if the test is 95% accurate, and 5% of a population is infected, then testing 100,000 people will result in 5,000 who either were told that they were sick but weren’t, or were told that they were not sick and are/were

At this point we haven’t seen any serology tests accurate enough to use for a specific person. However, those levels of accuracy are good to enough for a population. For instance, if you wanted to know what percentage of Knoxvillians had been exposed to COVID-19 a serology test would work great, but it wouldn’t be very reliable on a per-person basis. [1]

The PCR tests are accurate but we don’t have enough supplies to test everyone. Also, they are snapshot of who is infected now. The serology tests, by contrast, can be done more widely (we have enough supplies) and tell us, broadly, who has been infected, ever. (Ever = up to this point, we don’t know how long COVID-19 antibodies will stick around as the virus itself has only been around for a few months).

There are other tests being thrown around, saliva based tests, antigen tests, but we haven’t seen any that are far enough along to report on, much less be useful in the real world. Eventually, you can expect to see a combination of PCR and serological tests being used to help inform the policy response and understand where we are at with COVID-19.

Bonus points: The best epidemiological models for COVID-19, to date, appear to be based on SEIR: Susceptible (people), Exposed (people), Infectious (people) and Recovered (people, which includes those who died). PCR can tell you about Infected, serological can tell you about Susceptible.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from JHU, NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily.

[1] We are simplifying. This article goes over the accuracy in more detail and proper terminology. There are serological tests claiming higher accuracy but we haven’t seen any that have passed through peer review at those higher levels of accuracy.

Today we’re going to go over a question we got today, which will be our focus.

If you want to see cases and deaths see yesterday’s post, there are not material changes.

How does increased testing affect New Cases?

We had one question come in asking how increased testing would affect new cases. The expectation by public health experts and JM Addington is that increased testing will lead to an increased number of confirmed cases. In fact, the Knox County Health Department is considering pulling back from some of their events because they aren’t finding enough new cases.

There are two primary reasons that we test in public health:

  1. To find out if a specific individual is sick. In the case of COVID-19 this would be to ask them and anybody that they have been in contact with to self-isolate to prevent further spread
  2. To find out how widespread a disease is in a community or geography (city, county, state, prison, hospital, nursing home, etc.)

Clearly if you test more you are going to find more so the question becomes how much testing is enough and at what point does your level of concern go up?

One of the simplest ways is to look at the positivity rate, the number of total cases divided by total tests. In short, the positivity rate will give you an idea of (1) if you are doing enough testing, and (2) if the disease is spreading regardless of the number of confirmed cases. Both Dr. Fauci and Dr. Gottlieb have said the that US should shoot for a 10% rate. (See this article for a more in-depth discussion)

The typical way to calculate this is to take total cases, ever and divide by total tests, ever. For Tennessee, we get a 6.26% positivity rate.

We’re going to admit to playing armchair epidemiologist here and show you how we look at it, which is to look at the last 7 days of new cases and divide by the last 7 days of tests, which gives Tennessee a 2.03% positivity rate.

The US has a total positivity rate of 15.84% and 7-day positivity rate of 10.85%.

The first metric — total cases/test — will bias your rate towards whenever you had the most cases and the most testing, so you may end up measuring more where you were than where you’ve been. Taking a more recent set of data will tell you about where you are now.

We’re going to show you this data for the last 3 weeks, then we’ll talk about using it in conjunction with two other important metrics.

TN COVID-19 Positivity Rate, May 5th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

Note the y-axis changes scale here

TN COVID-19 Positivity Rate, May 5th, 2020 by JM Addington Technology Solutions
Source: Data from The COVID Tracking Project; Chart by JM Addington

The relatively low positivity rate in Tennessee is one of the reasons we’ve said that the data in Tennessee supports re-opening.

Of course, the positivity rate can be gamed: if you test a bunch of people that are not sick it will go down. The real question is how is the positivity rate moving with the total number of tests, the number of new cases and the number of new deaths?

You must combine data points to to look for the trend. These data points are cardinal directions, north vs south, not specific headings, 357′ vs 356′.

If cases are going up but the rate and deaths are stable or declining then it is likely you are simply doing a good job testing and the situation is stable or improving. If the rate is high, and stable or increasing then an outbreak may be getting ahead of you. If deaths, cases and positivity rate all increasing you’re in a world of hurt.

In a perfect world you will see cases, deaths and positivity rate all trend down as the situation improves. Nationally, the positivity rate (see above) is dropping faster than new cases or new deaths per day, which would be indicative that our testing is getting better but the situation is still static overall.

US COVID-19 new Cases & Deaths by Day
Source: Data from The COVID Tracking Project; Chart by JM Addington

In Tennessee, you see news cases per day jump mid-April and then plateau to where we are today, along with deaths. With a rising positivity rate this would be indicative of the situation worsening overall, if only slightly at this point.

TN COVID-19 new Cases & Deaths by Day
Source: Data from The COVID Tracking Project; Chart by JM Addington

…but…

Even all of these numbers together don’t do a good job at telling the story anymore. If anything, they are just enough to give you an arrow indicating direction.

Tennessee is testing all prison populations this week and already began testing all long term care facilities, both populations that have a lot of cases. So your tests and your positivity rate are going to go up because of the more targeted tests. This doesn’t mean that community-wide transmission has changed.

For better or worse, we think that this is going to increasingly be the case across the nation. Even the most in-depth numbers you have on testing, cases and deaths will only tell you part of the story. We’re not sure how keeping up with the numbers alone can be useful.

News

Building on yesterday’s news of a patient in France retrospectively being tested with COVID-19 on December 27th, 2019, is now bolstered by a peer-reviewed study showing that COVID-19 was circulating widely earlier than initially believed.

Tennessee is finding large amounts of asymptomatic cases in its prisons. Public Health Commissioner Dr. Piercey acknowledged as much during the briefing today but didn’t add any color to it. Combined with yesterday’s story of similar testing in a meatpacking plant, it challenges some of what we believe to be true about COVID-19.

Zoo Knoxville is set to re-open May 18th.

A Knoxville firm, one of Tom Boyd’s, is working on COVID-19 antibody testing.

Governor Bill Lee announced that Tennessee has procured 5,000,000 cloth masks and will be distributing them, free, via public health departments (from today’s briefing).

Infrared thermometers are available for Tennesseans only at tennesseepledge.com, $33/ea.

Dentists are allowed to open in Tennessee’s 89 rural counties, tomorrow.

By the way, if you want to see the same graphs for your state click here, we pull nearly all of our own visuals from this report. Typically updated at the same time the evening’s post goes up.

Here are GA new cases.

Also, you can do it on your phone but it is much easier to navigate on a bigger screen.

These posts helpful?

Then go ahead and share them where you saw them once or twice a week.

Get In Touch

Need help thinking this through? Access to more data? Help getting your technology in order to handle what’s here and what’s coming? Contact us today.

Other

Thanks to those that have shared kind words or liked these posts. We’re doing out best to put out data-driven analysis, each one of these takes about 2 hours. It’s helpful to know that they are being read.

If you want copies of the Excel sheet and PowerBI Reports we use to put these together email us, info@jmaddington.com. Right now, our PowerBI combines data from JHU, NYT, COVID Tracking Project, IMHE and the TN Department of Health. Most sets are updated daily.

[1] We are simplifying. This article goes over the accuracy in more detail and proper terminology. There are serological tests claiming higher accuracy but we haven’t seen any that have passed through peer review at those higher levels of accuracy.