Covid19 is escalating out of control in the US
Intro
In this short work you’ll see an analysis of death rate vs case rate from covid19. You’ll see that because death count lags case count, the US is almost certainly in for an alarming increase in deaths per 100K population due to cases already logged. It is nearly ordained that covid19 will be the #1 killer in the US by the start of Dec. And if case counts do not level off immediately, there will an even more alarming increase in death rate per 100K population. The US needs, as a population, to use masks and observe physical distancing.
Also, I’ve developed a metric “death rate per lagged new case rate”, which appears to have stabilized at just under 2% of those tested positive for covid19, which can be used for predictions going forward until a vaccine is in place.
Discussion
The US has for a long time been the #1 world leader of covid19 cases and deaths. In a recent work Covid19 spread in the USA — liberty vs science, state by state I discussed the state by state differences in covid19, discussing liberty vs science and observing the large difference in responses per state. A key conclusion was that states which favored liberty tended to be afflicted with more covid19 cases and deaths, and those which favored science tended to have fewer cases and deaths.
At this point, though, throught the US, the problem has become one of literally epidemic proportion. Unfortunately, with analogy to the old parable, those who favor liberty are drilling under their own seat in the combined boat, with the result that the boat itself is sinking.
Fig 1 shows the aggregate death rate per 100K population, calculated by myself for 7 day average sourced from primary data from Our World in Data, by annualizing data from each day and comparing to the entire US population as aggregate. A few features are clear. First, covid19 is already on the death rate map next to the US three top killers: heart disease, cancer and accidents (2018 data from was the CDC was used and converted as described previously) — covid19 is presently the nation’s #3 killer and increasing in rate. Next, the death rate peaked on or near Apr 21 and dropped significantly. At no time since then has the death rate reapproached this rate. This is falsely interpreted as “it’s just a flu; there is no problem”. In fact, the science predicts there is a powder keg in the making, as will be discussed below.
Fig 2 shows the mortality rate vs time, directly copied from Worldometers, a source of primary data. This means the mortality rate for known and resolved cases, not the population as a whole. It is very clear that the mortality rate from resolved cases, at the peak of the New York crisis, was running at some points in excess of 50%, and since that peak in April at which point in time the morality rate was still in excess of 30% of resolved cases, the aggregate mortality rate has declined to the 3.5% it is today. The present mortality rate may be more like 3% or so, since that 3.5% includes all the high mortality from the spring (I have not put the effort into doing this detailed calculation). Since the horrific spring wave, there are fewer deaths per 100K of population, even including the second “wave” which peaked near Aug 5, primarily because the mortality rate of closed cases has declined by a factor of more than 10x. There are two primary causes of lower mortality per contracted covid case: (1) we understand better how to prepare for and treat covid19 and (2) the virus itself mutates, and always favors continued propagation – if the disease is somewhat less fatal, there is less inhibition to the human interaction which spreads the disease, which promotes spreading the disease. The overall aggregate deaths in spring were confined to the northeast US when we didn’t understand the disease as well as we did today, with a mutation which arrived from Europe https://ltarof.medium.com/covid19-spread-in-the-usa-liberty-vs-science-819a2a6d8c2. However, the perceived benefit of lower mortality rate has run its course, because at this point in time, the full population of the US is involved, and spreading faster than ever, as will be discussed below.
Fig 3 shows the new case rate per 100K vs time, calculated by myself for 7 day average, also sourced from primary data from Our World in Data. Note that both new cases and deaths are 7-day averaged, because there is a dependence throughout the week on reporting – less cases are logged over the weekends. We can see there is a peak near Apr 12 in new cases, corresponding to a peak in death rate near Apr 21. The shape of the curve is different because of the dramatic improvements in mortality during that time (Fig 2). The “second wave” had a new case peak near Jul 24 correspond to a death peak near Aug 5. The shape of the curve is similar in both new case rate and death rate because the mortality rate was not changing as much — 10% at the start of July and 6% at the end of July. The shapes cannot be expected to be perfect because different parts of the US were increasing in new case count at different time – this is a blend rather than one geographic region. Yet is quite clear that deaths lag new cases, and predictably so. This is a good example of correlation with causation.
It’s worth understanding that Fig 2 is about mortality from known cases, and that Fig 1 is about mortality. To first order, Fig 1 is the convolution of Fig 2 and Fig 3. Or put another way, Fig 1 = Fig 2 x Fig 3 to first order.
Now look carefully at the rise in new case rate beginning near Sept 29 in Fig 3. A corresponding rise in death rate near Oct 17 in Fig 1 is observed. The death rate increase from Oct 1 to Nov 15, the last date recorded in this work, is in the ratio 1.76 – i.e. the Nov 15 death rate is 1.76x the Oct 17 death rate. The new case rate increase from Sept 29 for the same time interval — i.e. from Sept 29 to Oct 28, was a factor of 1.75x – this is the same number. We clearly observe that for the present wave at the time of this writing, the death rate lags the new case rate by approximately the interval from Sept 29 to Oct 17 — i.e 18 days.
Now look at the new case rate increase from Oct 28 until Nov 15, this same time interval of 18 days. This is in a ratio 2.06. This means that for the cases already logged by Nov 15, a death rate vs case rate lag analysis would predict a death rate approximately 2.06x what it is Nov 15, by Dec 3. This factor of 2x, shown in Fig 3 for new case rate, is also shown in Fig 1 for death rate. Put another way, the new case rate double already observed from Oct 28 to Nov 15 appears to preordain a doubling of the death rate from the Nov 15 point by approximately Dec 3. Most likely the aggregate US death rate from covid19 will be the nation’s #1 killer sometime before the start of Dec. As further evidence, consider that hospitals are already near capacity in many parts of the US. This is the price of so-called liberty – “live free and die”. And it gets even worse. It is clear from the trajectory in Fig 3 that the case count is increasing even further and faster. Even if everyone started using masks and isolating as of today, there are latent cases waiting to be identified and logged from spreading which has occurred until this time, which will also translate into more deaths.
Fig 4 takes this 18 day lag and calculates the death rate per new cases with an 18 day lag, as we learned from Figs 1 and 3. Note that the shape is quite similar to the orange curve in Fig 2. The orange curve in Fig 2 is aggregate mortality over all time prior to the date in question. This curve in Fig 4 is as close a proxy as I can calculate, at this time, for the death rate for new cases 18 days prior. Note that this rate appears to have stabilized at just below 2% since mid-July, and also gives additional confidence in the projection in Fig 1 based on Fig 3 using a constant death rate/lagged case rate, uncomfortable though the result is. Also, this is consistent with the orange curve in Fig 2 reducing in what looks like an asymptotic way — that asymptote appears to be just under 2%. I’ve personally not seen this calculation before. This suggests that going forward, the involuntary herd immunity would be at just under 2% of the infected population. In that case, I would down-revise the death toll for 80% of the US population to 5.2M, not the roughly 8M people as I recently estimated. Still, this is a massive death toll.
A bad as this is, we can, though, take control today and make this no worse than is already ordained. We all need to wear masks and physically distance whenever practical, to stop the spread. Otherwise, what already looks really bad today and will look even worse by the start of Dec, 2020 will become even worse that that, and has the potential to escalate into involuntary herd immunity. The 0.25M deaths as of Nov 15 is bad enough. Now think about many millions. And most of these deaths are preventable. Let’s all do our part, so we can reach the point where a vaccine can become effective.
Conclusion
Covid19 is poised to be the #1 killer in the US. The US needs a national mask mandate today. The US needs physical distancing today. The virus does not care about anyone’s liberty. It only cares about propagating from one person to the next. Please consider not drilling under your seat — everyone is in the same boat.
