How are you holding up?
I figured today was a good day to write another contemporaneous "journal" entry. These Weekenders have helped me keep my thoughts straight, plus I've been getting fantastic feedback and anecdotes from you awesome subscribers.
(There are 9,300 of you now!)
When this is all over, I hope to go back and see where I was right and wrong on 2019-nCoV and think through the whys.
So, here goes – another reflection on the state of COVID-19... biased towards the United States.
Today's Situation: 4/26/2020
As I write this today, 4/26/2020, pretty much every corner of the globe has its own local outbreak of COVID-19 to track. From a glance at Worldometer, the world at large now has 3,000,000 confirmed positive tests, and sadly 207,000 people thus far have died from the virus.
Here in the US it's also dire.
As of now, the United States has unfortunately identified 55,000 deaths to COVID-19. For those still holding onto the "just the flu" hypothesis, that puts it in line with the modeled number from the 2018 flu season of 61,000 deaths:
We'll undoubtedly pass 61,000 deaths – the next two substantial flu benchmarks are 1968/H3N2 at 100,000 deaths in the US and 1957-58/H2N2 with 116,000. Scaled to today's population, those pandemics would have killed 165,000 and 225,000, respectively.
Grim numbers – but there are some reasons to believe we may have passed the peak. In the next few weeks, we'll see some states move to reopen. Unfortunately, we'll see some false starts and second peaks too.
How Prevalent is COVID-19 in the United States?
The United States is currently testing around 200,000 people a day and finding 10-15% positive of late. But you're probably – kind of like me – wondering how prevalent COVID-19 actually is in the United States.
The 1,000,000 positive cases are only a lower bound (yeah, yeah, false positives – humor me. More than a million people here caught it.)
To that end, we're finally getting our eyes on preliminary serological and antibody tests from the field. These tests lack the accuracy of the swab tests for active infection, but they can pick up people who once had COVID-19 and got over it (or even never knew they were sick). Read more about them here.
Note some of these links are preprints or have poor design, but please get a feel for the _ranges_ folks:
- Santa Clara County, first Week of April 2.49% - 4.16% of people, ~50-85x undercount
- Chelsea, MA 33% of people(!), ~10x undercount
- Robbio, Italy 20% of people, ~10x undercount
- Castiglione d'Adda 66.6% of blood donors, Italy testing was overwhelmed some time back though
- Los Angeles, CA 2.8% - 5.6% of people, ~28x - 55x undercount
- New York: ~10x-11x undercount
- New York City: 21% of people
- Long Island: 16.7% of people
- Westchester and Rockland: 11.7% of people
- Rest of New York State: 3.6%
When Metros Get You Down
Here in the US it's popular in the media to look at the aggregate data. However, it's not the best way – I built this site on that.
Just as we track income by city here, things are vastly different in various parts of the country. And there are other factors in cities where there are no perfect comparisons – climate, latitude (sunlight!), who lives there, density, etc. Comparing the US to Italy is silly, and you should actually compare the US to... the EU. (But I digress)
If you allow there are ~300 metro areas in the United States, the virus is going to spread and morph in different ways in each one. But, as I felt compelled to complain about last time, sometimes you need to reason off incomplete data with huge error bars.
Some logical leaps are more logical than others, and all that.
I assume – despite the WHO's weird wording – that having antibodies means you'll be at least somewhat protected from reinfection. With that in mind, we're probably looking at something like this as of the end of April:
- 331 million people in the US
- 1 million with confirmed infections (.3%)
- 10x-50x undercount
- 10 million low prevalence (3%)
- 17.5 million best guess (5.3%)
- 50 million high prevalence (doubt this, but 15.1%)
They Call it the Herd...
Undershooting our official tested count is good for some of our measures yet bad for others. For one, a ton of undiagnosed cases implies the death rate of COVID-19 is well under our initial fears of the 2-4% ballpark. Thankfully.
But hold that thought. ~.13% of all of New York City has died in the last couple months – so the death rate is more than a terrible flu season's .1% (unless the dynamics of NYC are wildly different than everywhere else). In fact, it's probably much more – unless you think NYC is at herd immunity (more on that in a second.)
That's mixed but okay news! Yes. But... it also means our R0 estimate was wrong.
A few editions back, I told you somewhere in the low to mid-2s was a likely R0. If the prevalence is this high, it's probably a natural 3, 4, or 5+ R0. (Remember: human behavior can lower R0 with social distance, masks, etc.).
Herd immunity is, essentially, blocking a virus's spread by making most targets immune in some way. When enough people have immunity – either from fighting off the particular virus and gaining antibodies or memory, receiving a vaccine, or even inheriting antibodies – a virus will fizzle out (at least until there are no new clusters of vulnerable people to blast through).
Herd immunity doesn't mean people can't get it again. Just see: the periodic outbreaks of measles, mumps, etc. But it does mean the end of wild, unchecked spread.
The formula provided to us by epidemiologists for herd immunity is:
So, if R0 is 2.3:
~57% of people need to get it for herd immunity if the virus has an R0 of 2.3.
But if R0 is 5.0 instead?
Now, ~80% need to get it (!)
One massive caveat: this formula is shorthand. It doesn't mean the virus will stop circulating, and it doesn't mean behavioral changes won't also affect things, nor climate, nor something like test and trace, etc., etc.
But – and perhaps most importantly – it approximates one person as equally contagious or vulnerable as everyone else.
That's useful in a model, but you know it's not true – just as a Twitter account with a million followers drives more traffic, viruses get their boosts from super spreaders. South Korea saw that where one member of the Shincheonji Church of Jesus led to 700 people being infected(!). 85 cases in Greenwich, CT can be traced to a single party.
What that number is really saying, is a virus spread randomly needs roughly x% of interactions cut off.
Just like the angry one follower Twitter account tweeting away without doing much, a person who unfortunately gets sick but never really goes out anyway is different than a highly social person gaining immunity. A politician, DJ, religious leader, or the like may have 1,000s more – or 10,000s more – interactions than another person.
So – keep that in mind when reading these models. R0 itself can fall depending on our good and bad luck as the virus evolves. And to pound it home again – it'll be different per metro area/virus cluster (think of the flu "going around" a workplace when another is spared, same idea writ large).
On Sunlight, Immunity, and n=1
One thing you can do to improve your immunity is losing weight if you're overweight. There's some evidence (ahem, in rats) ketogenic diets are an even bigger bonus – but just do what you can where you can. If you don't work out often, check our BMI percentile calculator for more. If you do lift, the body fat percentile calculator is more appropriate.
Now, back to the (quicker) topic at hand.
Since I started musing here, I've been hot on Vitamin D and Zinc as supplements to help your own immunity. Don't get me wrong: I'm still hot on them. But, ridiculously, they've both now been politicized.
Zinc, as you know, was swept up in the hydroxychloroquine/chloroquine furor.
HCQ and its relatives (including regular old quinine in your tonic water) are zinc ionophores, shuttling zinc into cells where it can't easily reach. My bet today is another zinc ionophore like resveratrol, EGCG, quercetin, etc. is a safer bet than HCQ.
Those are just my (non-Doctor) thoughts today – don't drink fish cleaner, damnit. Zinc is the difference-maker you want, in my opinion. The ionophore you use is a distraction – and I bet zinc will turn out to be most useful before infection, or early in the course of COVID-19 infection.
Vitamin D is now suddenly a culture war supplement, too, after President Trump talked about disinfectants and sunlight.
But – culture war be damned – I'm even hotter on the sun now than I was a few weeks back (and you better believe my puns are intended). Here's the ammo (mostly preprints, some letters):
- Vitamin D Supplementation Could Possibly Improve Clinical Outcomes of Patients Infected with Coronavirus-2019
- Evidence that Vitamin D Supplementation Could Reduce Risk of Influenza and COVID-19 Infections and Deaths
- The Role of Vitamin D in Suppressing Cytokine Storm in COVID-19 Patients and Associated Mortality
- Vitamin D supplementation to prevent acute respiratory tract infections: systematic review and meta-analysis of individual participant data
- Vitamin D deficiency in Ireland – implications for COVID-19. Results from The Irish Longitudinal Study on Ageing
Checkbox 1: I'm 99.9989% sure Vitamin D interplays with whatever upper respiratory illnesses you pick up. Please don't avoid supplementing or getting sun because of this issue being politicized.
Now, on the sun and UV lights:
- Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1 SARS-CoV-2 survives for days on most non-copper surfaces
- Could ultraviolet lamps slow the spread of flu?
- Shanghai introduces ultraviolet light to disinfect public buses
- Indoor transmission of SARS-CoV-2
- (A throwback!) The Open-Air Treatment of PANDEMIC INFLUENZA
Checkbox 2: I'm 97% sure the dynamics of "outdoors" reduce the risk of the spread of COVID-19. It does for almost every other virus, as far as I know. And using UV-C to sanitize things definitely works.
But: we're talking about the outdoors.
I'll say it now: we will likely learn the subway in New York was a massive driver of infections. I would not be shocked if thousands of deaths were due to the subway. Not New Yorkers in parks, not New Yorkers greeting hospital ships, and not students returning to New York from spring break in Florida. I mean – the airplane air is probably more dangerous than the beach air.
But where are the lecturing stories about the subways? (I'll admit spring break is fun to blame, though. It's the avocado toast of annoying COVID anecdotes.)
Sun-power honestly just feels like one of those "stop insisting on a new RCT every time a virus mutates" conversations like masks was until 3-4 weeks back. (And for those of you who took exception to this to my last post – I'm not the only one dragging the media and this idea in general, please also read Slate Star Codex's take.)
Now, be safe – obviously, overexposure to the sun can cause all sorts of problems up to and including skin cancers. Vitamin D over-usage can cause hypercalcemia (it takes a ton though). You shouldn't use UV-C lamps on your skin (and many wavelengths of UV-C can cause cataracts if you look too long).
But, I'm calling it now – in a few weeks people will infuriatingly say, "we always knew the sun & outdoors helped."
No, people didn't. (Or at least they're not speaking up now.) Here's my marker. And remember you first read my Vitamin D thoughts two months ago.
On Treatments & Vaccines
I haven't seen anything either too promising, or too depressing here.
Since the last time we talked, the two most hyped drugs – hydroxychloroquine and Remdesivir – have received both positive and negative news. Other drugs like ivermectin have moved to clinical trials. And on the vaccine front, read Derek Lowe's coverage – there are some good candidates.
I'm hopeful here that we have a decently powerful treatment by fall. Either a prophylactic we can give to our front lines, or a treatment to knock down the death rate a bunch (convalescent plasma in quantity qualifies).
Vaccines are very iffy, just due to the timeline – even if we nail every trial, maybe there's a candidate in the early spring. To that end, please read about the folks who are volunteering to be exposed to COVID-19 to accelerate vaccine production. True heroes.
Stay Safe, Hope You're Doing Well!
This sucks, there's no getting around it.
In the Bay, we've been locked in the better part of two months now. Yes, with two daughters. (They feel the absurdity of the current situation, and remind us constantly.)
We do need to think about how to let more and more of the economy come back. Remember, not everyone can work from home – as I do, and you might be doing.
Initial claims have been depressing every week, and I have no doubt 20% of workers are currently in stasis. And the human toll has been devastating – we've lost 55,000 fellow Americans (and counting).
But let's end on a high note. There are reasons to be optimistic!
We've avoided the absolute worst predictions and model implications. This isn't the plague's death rate, folks. And the US's unemployment boost is low-key the best stimulus so far (ignore "but it incentivizes people not to work!" arguments – that's the damn point.)
Again, we're going to beat it. Expect successful reopenings, staged openings, and even some false starts soon.
But always bet on humans. We keep winning.