Coronavirus : Update 20 from the Epicenter

Posted by in Coronavirus

It’s been more than 3 months since Update 19, and your experience over that time will depend on where you live.

At the risk of generalizing, if you are in a relatively small European country with a strong central government that shut-down effectively then it’s likely that high infection rates were limited to one or two initial areas (Lombardy, Catalonia, Greater London), so you will have seen daily case numbers fall by >95% and stay low for several months :

Italy Daily Case numbers from mid-March to mid-August

Whereas if you live in a large country with a devolved power structure (i.e. strong states) then it’s likely that lock-downs were less effective and you will have seen outbreaks replicate across the country (e.g. NY, TX, FL, CA…) :

U.S. Daily Case numbers from mid-March to mid-August

So, economic factors aside, the European lock-downs and travel restrictions appeared to work, but as restrictions were lifted and vacation hotspots began receiving (mostly domestic) visitors, so those enviably low infection rates started rising again, with Spain and France suffering particularly large spikes. In the chart below we can see the steady rise of the large countries on the right contrasted against the s-curves of the Europeans :

Total Cases for select European countries (left) vs large countries (right). Note that the right scale is 10x the left scale

This difference is also evident in the Daily Case numbers :

Daily Case numbers for select European countries (left) vs large countries (rights). Note that the right scale is 5x the left scale

So it seems that countries have to chose between a series of stop-start interventions or a slow-burn whose speed is determined by the ability of a country’s citizens to absorb the long-term cultural and financial impact of social distancing.

The third model – a push for all-out herd immunity that Sweden has so controversially pioneered, may prove in the long-run to be a winning bet. In the chart below we look at the three approaches as followed by France, the U.S. and Sweden – plotted per million inhabitants. Lock-downs appear to work as long as you stay locked down; slow-burns lead to high plateaus, and herd immunity permanently exhausts the supply of people to infect :

France vs U.S. vs Sweden (daily cases per million)

Well, this isn’t really news. We discussed exactly these ideas in previous updates. Way back in mid-March Update 6 examined Imperial College’s start-stop hypothesis, that matches the UK’s situation remarkably well :

Imperial College phased-lockdown predictions vs UK reality

What’s dramatically different today compared to March is the rate at which the infected are dying. At the beginning of the epidemic, we used case numbers as a 14-day predictor of death rates – due to the strong correlation between the two numbers :

France Daily Cases (blue) vs Daily Deaths (red) – February to June

Something like 20% of the total infected population were dying during the first wave, and so minimizing infection rates was critical to avoid swamping ICUs, exhausting ventilator supplies, and spiking death tolls.

But what is really important to understand is that despite the aggression of the second wave of infections, the death rates have not climbed proportionally. Here is the same chart from France, extended to mid-September :

France Daily Cases (blue) vs Daily Deaths (red) – February to September

Woah! Crazy Second Wave of infections! But look at the red curve. Very few people are dying. This is a ‘safer’ wave compared to the first in most countries. We can see how this phenomenon plays out across multiple countries by plotting the ratio of deaths to cases :

Death to Case ratios per country

Remember that testing rates has a big impact on these numbers. Testing has increased significantly over this time-frame and so we’re catching many more asymptomatic cases than we were in March, but you only have to look at the absolute death numbers to see that the majority of countries seem to have ‘cracked the code’. Clear exceptions are the U.S. (whose numbers have fallen, but remain at a higher than average plateau – mainly due to persistently poor performances in CA, TX and FL) and Brazil (whose steadily increasing numbers appear to strangely mirror those of the U.S.) :

Leaving the outliers aside for a moment, let’s look at what might account for the sudden drop in mortality. I have broken it down into 3 main areas :

  1. Lower average age of infected population
  2. Improved medical understanding / procedures / drugs
  3. Sunlight & Vitamin D
  4. Higher humidity

I will follow up with a post about each of these, but we can summarize :

More testing across the board, more risk awareness and precautions taken by the elderly, and the opposite over the summer from the young has shifted the average age of infections sharply down. As Case Fatality Rates drop by age, so we have seen both a fall in mortality and absolute fatality numbers.

As reported in Update 15, the medical community was aware as far back as mid-April that the virus did more than just affect lung function. Inflammation, and blood clotting was known to be the underlying problem, and while the exact mechanisms were still unclear, they knew that intubating people at the first sign of breathing difficulties was doing (much more) harm than good. Anti-inflammatory drugs, anti-coagulants, and even arthritis medication could stop the body’s own immune system from going haywire and causing damage to the lungs, heart and brain. More recent research has pinpointed the mechanisms involved more clearly, and has identified already available drugs that can calm the so-called bradykinin storm that the virus causes in some patients. Except for the most elderly and infirm, this all means that you’re now far less likely to die of a Covid infection than in March.

At the peak of the epidemic, researches noticed that 86% of Covid victims were vitamin D deficient. Vitamin D it is actually a hormone, and its effect on the body is multifaceted. For Covid sufferers the most important function is that of immune regulator – helping protect against runaway immune responses. The body produces vitamin D naturally when exposed to sunlight. 15 minutes a day is enough, and look what happened to fatality rates as we exited lockdown and enjoyed the feel of sunlight on our faces. Key takeaways here are that high-melanin skin reduces the production rate of the vitamin, which is why people of color living in high-latitude cities may have higher Covid mortality rates. The good news is you can take a pill. 5,000 iu / day is plenty.

The most important form of viral transmission appears to be airborne, which is why wearing a mask is important. In humid places droplets formed by shouting, singing, coughing or sneezing grow quickly in weight and fall to the floor. In dryer conditions the virus can stay in the air for hours and cause mega-spreading events. Dry air also makes the mucus membranes in our noses dryer and more permeable.

Public buildings should have at least 40% humidity indoors, capped at 60%, to reduce viral-spread risks for occupants

Sumit Kumar Mishra  of India’s CSIR National Physical Laboratory and Alfred Wiedensohler and Ajit Ahlawat of Germany’s Leibniz Institute for Tropospheric Research

With winter coming in the northern hemisphere, we’re going to have a nasty combination of lower sunlight, more indoor gatherings, lower humidity levels, and complicating factors like regular flu. It’s extremely probable that infection rates will soar, and ICUs will begin to fill up – prompting the WHO to push governments to lock-down their populations again in an attempt to keep case numbers at bay. But with millions now protected with antibodies, and mortality largely under control – would keeping billions of people from contributing to the economy be more a form of virtue signalling on a massive scale than a public health imperative? We’ll see how this unfolds…

All the charts shown are taken from my Covid-charts notebook that is replicated here on

For those that are interested, the title picture is of viral particles (virions) in the lungs of a patient in the grip of a bradykinin storm.