As Of Last Week, The US Has More Recovered COVID Cases Than Active Ones

We crossed an important milestone last week when the number of recovered COVID cases surpassed the number of active cases.  This hasn't been widely reported because the US doesn't have an official recovered cases count. 

Sites like Worldometers, where I pull much of my data, include a recovered cases count, but they admit this is a very imperfect count because of the inconsistency that different countries report their recovered cases.  In my last report, I graphed US recovered cases on the assumption that recovery is 21 days after symptom onset.   This also isn't perfect because some hospitalized cases can take much longer to recover, but also many people are tested and confirmed well after symptom onset an so it seems like a conservative number (NC has started reporting recoveries using a 14 day estimate). 

If you subtract the recovered cases and fatalities from the current case count, you can also get an active case count.  Here's graph showing the four numbers:

Our active cases as of today are 38% of our total cases reported -- a little more than 1/3.   For comparison, I added a recovery calculation and % active cases to my current data table. The % active cases gives a good read on the progress a country has made in the pandemic. Here's a summary:

Country	Days post 100 cases	Calculated Recoveries using 21 day period	Calculated Active Cases	% Active Cases
USA	73	801,245	543,906	38.03%
Italy	81	158,867	32,131	14.47%
Spain	73	158,867	32,131	21.42%
Germany	74	10,443	260	12.04%
France	75	185,920	58,071	10.24%
UK	70	145,268	20,969	39.89%
Canada	62	132,754	18,232	41.74%
Sweden	69	104,892	91,627	41.97%
India	62	36,808	30,168	70.48%
Indonesia	59	13,315	12,134	49.64%
Philippines	61	20,488	55,016	39.91%
Vietnam	53	6,747	7,663	6.94%
Thailand	59	6,209	4,637	5.90%
Malaysia	66	268	20	17.35%
Singapore	73	2,783	178	55.90%
South Africa	58	5,494	1,176	67.26%
Argentina	56	11,157	14,168	50.07%
Australia	64	3,734	8,121	4.48%
Chile	59	3,106	3,444	65.64%
New Zealand	53	6,569	313	3.07%

I hope to post a follow up with recovery data for NC soon. 

May 13 - Some Positive COVID Trends For the US and NC

The increase in testing over the last couple weeks and the effect of preventative measures on the spread of COVID has resulted in some pretty promising trends lately.

First, here's some notes on our current status: 

• Our 4255 cases per 1M people ranks us at number 3 in our 22 country sample (Singapore and Spain are the two with higher rates)
• Our 252 fatalities per 1M ranks us 6th
• We have now tested over 29,000 tests per million people for COVID - almost 3% of our population.  While other countries have higher per-capita numbers,  our 9.6 Million overall tests is more than 3 times the next closest country (Germany). 
• We are now at 25 days to double and that number is going up every one or two days. On April 1, we were doubling in 4 days. 
• We are 18 days out from our peak case day of 38,790
• We are 26 days out from our peak deaths day of 6,088

While this graph is a little chaotic, it shows how most countries positive test rates for COVID are going down:

Singapore's rate is the one that is obviously climbing.  Canada is just leveling out.  You can see that the US test rate peaked several weeks ago and is on a good downward trend. 

When we look at regional growth,  you still see the tropical and southern countries very flat relative to the northern countries.

The two exceptions are Chile and Singapore.  About half of Chile's population lives in the area around Santiago, but I don't have an explanation for their anamalous growth compared to neighbor Argentina.  I noted last week that Singapore uses much more air conditioning than other tropical countries on this chart. That may be a factor, but there are certainly many other possibilities. 

Focusing on the US, I calculated the number of COVID cases that have recovered I used an average of 21 days to recovery that is the consensus of several studies.  If you subtract the recoveries and fatalities from the total cases, you can also see the number of active cases. 

From this graph, you can see that there are currently more people who have recovered from COVID in the US than active cases.  Also, the number of active cases slowed around a month ago and peaked on April 25th at 621,872.  Since then the number of active cases has been dropping. If we look at the daily case and fatality rates, we also see that our peaks are weeks behind us:

Finally,  our testing is headed in the right direction.  The number of tests per day has grown to three times its rate since the beginning of April while the % of positive cases continues to drop.

Looking at North Carolina,  

Our new cases per day remains the lowest among our peer set, although we have been increasing while others have been going down.  Some of that is because of the increased testing we've been doing, but these other states have also increased testing rates.   Our daily mortality rate remains one of the lowest per-capita in the country and its a good sign that the others are trending our way. 

By the way, all these graphs and the data behind them are updated daily.  The world graphs are here, the national graphs here, and the NC Graphs are here.  Here's a comparison of testing:

Florida has increased its rate, but they have twice our population so you would expect more testing.  Michigan would be a higher priority for testing because of their more severe COVID situation.  North Carolina is about 30% off the national pace and is testing at a lower pace than some of the other states, but to the extent these tests are prioritized at a national level, we have a relatively mild COVID situation so you would expect more tests to go to states in greater crisis.   When we look at the positive test rate, we also see that NC has always had a relatively low positive rate and that the other states are starting to catch up.

Looking at the combination of daily testing rate and positive testing, we see that the lighter trend lines are in the right directions (testing up, positive rate down).  This is one of the key indicators NC is using to determine if its okay to go to the next phase of reopenings. 

Finally,  I have also added NC county-by-county data to my COVID data set.  A sortable summary is here.   Here are some interesting facts:

• 89% of counties are 7 days or more from their peak case date
• 85% are 7 days or more from their peak death date
• 93% are more than 7 days to double
• 16% haven't had a new case in at least a week
• Charlotte's county, Mecklenburg, has almost 25% of the entire state's cases and more than twice the next closest state.  
• The NC Prison system has more cases than all but 4 counties and one of those counties (Wayne) has many of the prison cases.   

 

COVID May Have Created The Best Time Ever To Buy A Car

One of the under reported stories about COVID has been its potential impact on the auto industry.  This year could be the best time in a generation to buy a car.  

There has been the full stop of auto auctions.  Normally this is where rental companies, corporations with car fleets, and others dispose of their vehicles.  Car dealers around the country attend large scale auctions where they buy inventory of these used cars to sell on their lots.  There hasn't been an auction since mid-march when COVID set in, which has cars scheduled to rotate off of fleets piling up.  Large auction houses like Atlanta-based Cox automotive are trying to pivot to a digital auction program, but Cox furloughed 12,500 workers this week , which could meant they know this will be a long process.  Manheim, an auction subsidiary of Cox posted this chart showing how sales prices have dropped from April last year across all market categories:

Click to Enlarge

A VP of Cox recently wrote an open letter to dealers where he predicted that "The market isn’t likely to improve significantly in 30, 60 or even 90 days."  He goes on to lay out three facts of what will put additional pressure on dealers:

• In six months, every car on the lot will be a model year older.  One-to-three year old used vehicles are in a weak spot because the factories will likely provide help for the dealer to clear out their new cars, but the dealer is on their own for 2019 and earlier cars. 
• In six months, there will be huge levels of wholesale supply on the market as all the rental fleets and dealer-owned used vehicles come to auctions. 
• Unemployment and a sluggish economy will likely depress sales even further. 

When they are moving again,  there will be extreme pressure to move cars quickly to make room for the ones backed up in corporate fleets.  A few weeks ago, we saw oil futures go negative -- people were paying to take oil off their hands.  That situation has since corrected, but it was due to oil coming out of wells with no place to put it.  The auto industry is facing the same dynamic of used cars flowing into a market with few takers.  I'm not saying someone is going to pay you to take their car, but I would expect dealers to be considering deals they would never offer before to get cars off the lot. 

The pressure will likely increase into the summer as the 2021 models start arriving on lots and all the existing cars become another year older.  Auto makers will likely pull out all the stops to create incentive for customers to buy the 2020 models to make room for the new ones. Also, the cars coming off rental fleets will have spent 3-4 months parked on the lots going practically unused so they will be some of the lowest mileage cars to see auction once they make it there. 

While all these factors can make an attractive market for a car buyer,  it is a list of miseries for car dealers.  Most in this business probably came into 2020 expecting a great year with wages going up, record employment, and a high confidence in the economy.  In addition to the shock of the lock down, Many dealers probably took on more cars expecting great returns and now the interest on the debt they took to buy those cars is now draining their cash reserves. Many won't make it.  Often, car dealers are long term members of their communities,  giving to charity and sponsoring youth sports leagues.  If you do choose to take advantage of the market situation,  I encourage you to be kind about it.  

DISCLOSURE:  I've never worked in the auto business, but picked up some of the market dynamics from friends who do.  If you're close to the industry and see a flaw in my reasoning,  you're probably right and should post in the comments. 

As we begin to reopen, what are the high risk activities?

Yesterday, my friend Lucinda posted a well written article by Dr. Erin Bromage: The Risks - Know Them - Avoid Them.  I have some issues with some of his statements, but I strongly agree with his main thesis:  There's a certain number of viral particles someone needs to take in for them to be infected with COVID and you can either get them from one big burst (like someone coughing in your face) or in smaller doses over time (like having a 50 minute conversation with someone who never coughs or sneezes).  The result is that some activities, like shopping, may be less hazardous than you imagine and others, like eating in a half empty restaurant, may be much more hazardous.

Rather than just rehashing Erin's well-crafted explanation of this concept,  I encourage you to spend a few minutes reading it before I get into the implications of viral loads.


 All finished?  Great!  Let's dig in...

Implications Of Asymptomatic Carriers. 
By now,  everyone has probably heard of the large number of people who have tested positive for COVID but never experienced any symptoms -- if they hadn't been tested, they'd have no idea they had it.  There's a big question about whether these people can pass on the virus.  If they can, then it means that the strategy of self-quarantining people with COVID symptoms won't much slow the pandemic because of the many people who would have no known reasons to self quarantine that could continue to infect others.  For a long time, I didn't think this was a big factor because the original case tracking in Wuhan found very few, if any, cases that couldn't be connected with a know person with symptoms -- there were few if any cases where someone got the virus and nobody could identify someone with symptoms who could have given it to them.  The second reason is that if you're not coughing and sneezing,  how much viral material can you really be passing on to others.

How many people have been infected with COVID by asymptomatic transmission is still up to question, but the research cited in Erin's article shows that virus droplets can be passed just through breathing, talking, and other basic behaviors ( the research was on other diseases like the flu and tuberculosis, but it's reasonable to expect its also true of  COVID.  The transmission is at much lower amounts, though, so you're unlikely to get COVID just by making short contact with an asymptomatic, but if you spend a few hours (or share a home) with them,  your chances could be good.

Quantity May Affect Severity
It's been found in other viruses like the flu that the amount of the virus you initially take in effects the severity of your illness.  It hasn't been confirmed if this is the case with COVID, but the number of very light and asymptomatic cases makes it a plausible hypothesis.  Someone with a more severe COVID case would have a bad cough and other symptoms that could transmit a lot of virus in a brief time, but they would also (hopefully) self-quarantine soon after they started experiencing the symptoms so there wouldn't be many opportunities to infect others.  On the other hand, someone with little or no symptoms with potentially be in contact with more people over more time, but would be distributing much lower volumes of virus particles. From the articles example,  a single cough could potentially deliver many times the minimum dose to infect someone but it might take a 50 minute conversation with someone with a light case to get infected.  My suspicion is that those who have light or asymptomatic cases most often got the minimal dose for an infection from other asymptomatics, who barely gave them enough of the virus to be infected.

To be clear, there hasn't been enough studies to prove this suspicion, and I'm stacking one unknown (does amount of virus effect whether someone is asymptomatic) on another unknown (do asymptomatics infect others).   I should also point out that research has found that people tend to have the highest quantity of viral particles right around the time they start experiencing symptoms so most people with COVID could be infecting others with lower doses before they start coughing and sneezing.

If we believe that masks are effective at trapping most of the viral particles an infected person has (I haven't seen a study confirming it, but it makes intuitive sense), then wearing masks while around others would be a game changer in terms of reducing the amount of particles others are absorbing while around you.  While non N-95 masks may not filter something as small as a COVID particle, even if it trapped half the particles it would mean it would take 100 minutes of conversation with someone infected to become infected instead of 50 minutes.

I should also note that just because someone isn't coughing or sneezing doesn't mean they can't deliver a high-volume of virus particles to someone else.  If you are in the same home with someone else, you have hours and hours to infect them -- potentially at high levels.  That's what happened with measles.  Typically, one child would have a casual contact with someone infected and bring a light case of measles home with them. Then their family would experience more severe cases from longer exposure to that child.

If it turns out that people with light or no-symptom cases of COVID tend to infect others with similarly mild cases, it could be very good news because it may mean that COVID is self-inoculating -- along with the visible severe cases, it even more quickly spreads light cases that can lead to group immunity. For this scenario to happen, people with light cases would have to develop COVID immunity -- something that also hasn't been determined yet.

Besides severity and asymtomatic carriers, there are many more implications of Erin's essay I'd like to explore:

Air Conditioning:  It seems likely (and early research supports) that once outside the body, COVID does much better in dry cool air (and electric light)  than it does in warm humid air and sunlight.  I've pointed out that tropical countries have had much milder COVID experiences than northern ones (see Graph), but Singapore has recently had some concerning COVID growth.  Singapore's founder practically build the country on Air Conditioning while the other countries in Southeast Asia still operate largely without it (though they're increasing adoption).  I ran a comparison of AC use and COVID Cases and saw what at least looks like a correlation:

Note that the US has lower AC use per capita than Singapore because of the winter season, but during that time, we are also in cooler, dryer spaces.  If AC is a factor in COVID spread,  then some of the less effected states, like NC, may see a spike in July/August where people tend to retreat indoors to air conditioned spaces.

Affect on different businesses:
If the guesses above are right,  it leads to a few principles on what's (relatively) low risk and what's high risk:

LOWER RISK

• Staying Home (if you know those at home are not infected)
• Short Interactions with strangers
• Interactions outdoors with fresh air (and better yet Sun)
• Locations where most or all are wearing masks

HIGHER RISK

• Extended Time Indoors with the same strangers
• Venues where many people are talking
• Air Conditioned spaces with many others and recirculated air
• Places where groups are largely not masked
• Working in a place that has a lot of interactions with potentially infected people

So what does that suggest about the prospects of different places reopening? 

• Most shopping venues are relatively safe (for the shoppers), especially if masks are required in the shop. The shop clerks are at higher risk because of the number of potentially infected people they encounter.  The plexiglass screens should significantly reduce their exposure, though, because most of the particles will fall downwards when a person breathes. 
• Restaurants will be hard.  You can't wear masks while eating, so you would potentially have a lot of unfiltered particles.  Servers can't stand behind plexiglass and may have trouble wearing a mask since many people will have a hard time understanding them.  Outdoor seating, even on hot days, should be safer.  Shorter dining times should help for the patrons.  
• Bars, particularly indoor ones, involve a lot of people close together, unmasked, and potentially with impaired judgement. People can spend much longer at a bar than at a restaurant as well, increasing their exposure to anyone infected in the space.  Bartenders will have an even harder time.
• Hair and beauty places seem relatively safe (for the patrons) for procedures that are an hour or less.  Those working there are at higher risk.  
• Movie Theaters -- Air conditioned spaces with a number of strangers. If people can get comfortable wearing a mask for 2 hours, it be reasonably safe. Comedies and horror movies that provoke a lot of laughter and screams from the audience would make their situations riskier.  There will have to be a way to clean all the surfaces in a theater between shows, which will reduce the number of shows a movie can provide.  Also for seating,  groups who are already exposed to each other will be clustered, but away from other groups.  This could require a smart seating reservation system that automatically blocks out seats to space groups who book together. 
• Churches -- the higher risk time would be when everyone is singing together.  Singing normally involves taking deeper breaths and expelling more air than talking.  It would also be harder to do through masks.  Churches opening in the next couple weeks should consider playing recorded music and asking the participants to join silently in their hearts.   Communion and offerings could also be tricky because it normally involves people handing a basket or elements from one end of the row to the other, which doesn't leave room for cleaning surfaces. 
• Sporting Events -- outdoor events where there is air circulation and (sometimes) warm air and sunlight are less risky than indoor ones.  Masks should be worn, and teams should be working to provide something fun with team colors that can be given away at the door (and paid for through higher ticket prices). Several companies have already produced Banditos (or tube scarves) and you can expect Nike, Adidas, Under-Armor, and others to jump on this train quickly.  The high risk time here seems to be in an enclosed air-conditioned space roaring with fans.  Extended fan noise is like singing -- everyone expelling more from their lungs and taking deep breaths.  
• Office spaces --  workspaces will move more towards social distancing --less community tables and more cubicles and offices.  The increased space per worker may be offset by having workers rotate between the office and home.  Traditional air conditioned spaces with no open windows will seem less safe.  Closed meeting rooms is where the real focus will come because you will have people grouped together and all talking.  Buildings may need to focus on the air handling of these conference rooms to constantly move fresh air into the space while keeping the temperature comfortable.  
• Schools -- There was a promising report recently that looked at COVID contact tracing from several studies and found there wasn't a single case of a child under ten passing on the illness to an adult.  The study also reinforced that children tend to get much milder cases when they do contract COVID.  If the research holds up, it could make schools safer for students and teachers than was feared. On the other hand, some school districts are already proposing removing PE, Choir, Band, and other large group programs from schools.  

Clearly from these examples,  reopening plans will be complicated and almost every industry will have different challenges.  Eventually, vaccines, group immunity, and therapies to reduce the severity of COVID will reduce the risks and make all activities safer, but you can expect it to be slow moving and with many mistakes.  Best to be patient, considerate, and thoughtful about what risk you take in the meantime.