Updated: May 26
This is the fourth in a series of blog posts about the coronavirus.
There is no "spin" in science. So welcome to the no politician zone.
My promise to you is that I will quote and cite the best medical journals in the world (Lancet, JAMA and the New England Journal of Medicine) and the World Health Organization. If I have to quote the media, it will be clearly stated.
In this post, I'll address how severe the disease could get on a macro scale in America. I'll also briefly mention the cures and treatments in the pipeline.
Another Simple Statistic
Let's talk about CFR, or the case fatality rate. That’s the percentage of people who die after contracting the virus.
Comparisons to the Flu
Let’s consider a seasonal flu fatality rate of 0.1% (i.e., it kills 1 in 1000 people). (Note that the Lancet says 0.1% represents a "moderately" bad flu season while the New England Journal of Medicine calls 0.1% a "severe" one). It is true that this 0.1% leads to tens of thousands of annual flu deaths in the US. (These of course are mostly the elderly and the very young—groups who have compromised or underdeveloped immune systems).
There have been less than one hundred coronavirus deaths in America thus far (as of 3/13/20) and people are using the annual number of flu deaths in America, by itself, to imply that this virus is nothing to worry about.
It’s a logical fallacy to use that one stat as a basis to say "we certainly are not in harm’s way; don't do anything to prepare." Imagine a Frenchman in 1914 saying (after two hours of WW1 fighting and a hundred dead), "this won't be anything compared to our war with the Germans in 1870. We lost a 140,000 men in that one!" By 1918, nearly 1.4 million French soldiers had lost their lives.
The Coronavirus Fatality Rate
The fatality rate from coronavirus in China is 3.7%—and that’s based on nearly 80,000 cases there. (This number and all of the remaining severity numbers were taken from the Johns Hopkins Covid-19 tracker on 3/9/20).
Let’s make a few points about this number:
1) That means coronavirus in China is 37 times more lethal than a moderate to severe seasonal flu in the United States. Note that this is an absurd comparison. It’s not even apples to oranges. It’s apples to aardvarks. It’s a comparison of one disease in one country compared to another disease in another country, so it’s hard to get perspective from that. Nevertheless, it does give us some sense of just how big this 3.7% number is when you compare it to 0.1%.
2) My hope is that this high 3.7% rate is largely due to the Chinese inclination to smoke cigarettes. This is a respiratory disease after all, and our smoking rates are much lower than theirs. The WHO says they smoke at a 47.6% rate; we are at 19.5%.
3) The 3.7% number could be dramatically off for other reasons:
--It’s calculated by simply dividing the number of deaths by those diagnosed with the disease. But are we really capturing all of the people who actually have the disease? Surely, we are not. Many patients will experience mild symptoms, not seek treatment and never get diagnosed.
--However, there are still people who have not recovered from the disease and many of those have not died yet, so they aren't counted. Because of this, in the early weeks of China’s outbreak, their fatality rate was 1% and then 2%. It rose significantly over time and is now stabilizing. (Please see the side bar example to get into the math basics on these two issues).
In South Korea (7,513 cases), the fatality rate is only 0.7% which is the most hopeful stat I’ve seen—and they have a lot of cases.
(Note that this is only one half of the equation, however. If 2,000 people get a disease with a 0.7% fatality rate, 14 people have died. If 2 billion people get a disease with the same 0.7% fatality rate, 14 million people have died. That’s why disease spread, addressed in the previous post, is so important.)
In Italy (9,172 cases), the fatality rate is 5.0%.
In Iran (7,161), the fatality rate is 3.3%.
The big question for us is: what will the fatality rate be in the United States?
Let’s consider the range of possibilities:
The WHO estimated on February 19th that the fatality rate will range between 0.3% and 1%. Why is this number so low compared to these other values? This suggests (although they have not explicitly stated this to my knowledge) that the number of people who are actually infected is much, much higher than those who have been diagnosed. That is, they think the denominator is higher than we appreciate.
After an outbreak, it's easy to count the number of dead, but it's also possible to perform "serology." They take blood (serum) from a random sample of people to see who has immunity to the disease. That tells you how many citizens were actually infected. Then, you have an accurate numerator, denominator and thus, an accurate fatality rate. I suspect the WHO has used serology to arrive at the 0.3% - 1.0% estimate.
The WHO did not indicate an expected value inside this range. They only gave the range. For my analysis, I’ll assume it could be 0.65%, right in the middle.
Remember from my previous post on disease spread that The New England Journal of Medicine's best estimate for R-nought is 2.2, which implies 54% will acquire Covid-19.
We can now estimate how many will die in the United States. Note that there will be many caveats to this number further down.
This is how many Americans could die, according to the scientists' numbers:
330 million Americans X 54% infected X 0.65% fatality rate = 1.15 million deaths
However, the actual R-nought value lies between 1.4 and 3.9. This dictates that between 29 and 74% of Americans will get the virus.
Here is a possible range of values depending on the ranges of fatality rates and r-noughts:
Caveats and Context
1) This assumes that we don’t take any drastic measures in reducing the spread of the disease. So far, we have stopped flights from China and Europe (as of 3/11/20) and many companies have encouraged employees to work from home. Concerts and conferences have been cancelled. Lots of school districts have closed. That’s a good start, but I’m not sure it’s enough. China and South Korea had huge swaths of their population unable to leave their houses. According to one of my (Pearl Jam concert) buddies who lives in the (highly affected) Veneto Province of Italy, they can go to the grocery store and pharmacy. They are allowed to walk or run outside but they have to stay away from one another. They are allowed to go to work in an emergency. So while the lockdown there is not as extreme as it was in China and South Korea, it is certainly much more stringent than what we're doing. (I outlined some of the extreme measures that China had to take in this post).
2) The death totals above assume that the data from China is correct. To allow for wiggle room, scientists give ranges of numbers that are possible. They crunch a bunch of data and give a 95% confidence interval. (That's the R-nought range of 1.4 to 3.9 for example). They are basically saying: “we are 95% confident that the actual number lies in this range.” The numbers closer to the end of the range are much less likely than the ones in the middle of the range.
3) This assumes that the data from China applies to America. Let’s examine that.
--The contagiousness (R-nought) could be higher here. We might be more social and gather more often. That would increase our R-nought and the number of infected. We might be less obedient than they were when it comes to hand washing. That would increase our R-nought and the number of infected.
--The contagiousness (R-nought) could be lower here. We might experience much warmer temperatures than they did in January and February and the virus might not like that. That would lower R-nought and the number of infected. The Lancet says clearly that they don’t know if this will be the case. Don’t bet on this. Hope is not a strategy.
--The fatality rate could be higher here. If people ignore personal hygiene directives and hospitals are overrun with people, the 0.3% – 1.0% estimate may not apply to us.
--The fatality rate could be lower here. If smoking plays a huge role in whether or not people die, then we may be better off here.
4) How does this compare to other diseases? The CDC says that nearly 600,000 Americans die from cancer each year and that another 650,000 die from heart disease annually. So with a range between 287k and 2.4 million deaths, comparatively, this could be a major event.
If we all are led to believe that this is a hoax or we don’t take it seriously, then huge numbers of us will contract the disease and overwhelm our healthcare systems. Hospital beds and ICUs will be full and new patients needing help won’t benefit as much from our advanced healthcare system. Thus, our fatality rate could be higher (compared to South Korea for example where they are taking it very seriously).
It’s impossible to tell what is going to happen, but thus far, I don’t like what I’m seeing.
The Question That Everyone Is Asking
When will a vaccine be ready?
Gilead's antiviral drug Remdesivir is being tested in a few different trials which won't be complete until April according to the company.
There are two advantages to approving Remdesivir both related to its prior approval for Ebola. First, the company doesn’t have to prove again that it’s safe, so it skipped FDA Phase I and II. Gilead’s job in these Phase III trials is to prove that a) it can treat an already existing disease or b) it prevents the acquisition of this coronavirus just as the MMR vaccine prevents the acquisition of measles, mumps and rubella.
Second, since the drug has already been manufactured, Gilead presumably could scale drug manufacturing more quickly than normal.
Keep your fingers crossed that Remdesivir is approved. But remember, hope is not a strategy.
A totally new vaccine, if it's needed after a remdesivir failure, will take months just to complete initial testing. Then they would have to manufacture it on a wide scale, which takes time. I have read various projections on when a vaccine would be widely distributed and how many vaccines are being developed. Most say the 20 or so vaccine candidates are 18 months away (July of 2021), but the numbers are all over the map, so I won't bother trying to quote them here.
Note that there is no vaccine for the common cold, HIV, malaria, tuberculosis, West Nile virus, Zika virus, Lyme disease and Hep C. So, don't pin all of your hopes on a vaccine. Hope is not a strategy.
The Question That Everyone Should Be Asking
When will we all start acting responsibly?
The vaccine talk misses the point! We can end this within a month, together, by taking critical precautionary measures, which I'll address in the next post.
We also need to spread factual information and help each other out. If you appreciate what I've done and believe that fact-based information may help someone you love, please forward this to them. This may save a few lives.