Thursday, March 26, 2020

No, It Is NOT Great

I have moved most of my COVID-19 posts onto a separate blog: covid19science.blogspot.com/. But I leave any related posts that are somewhat political here, and I have to add one.

I usually do not watch the White House briefings, because I have a very low tolerance for lies and incompetence. But my friend Pierre, a mathematician, mentioned that Dr. Birx seemed to have some problems understanding the models, so I looked at that part on Youtube.

Unfortunately, she made factually wrong statements less than 2 minutes into her talk. The first one was "In no country have we seen an attack rate over 1 in a thousand". One look at the numbers for Italy (and perhaps a bit of help from a calculator) would have shown her this is wrong. Italy has 80,589 cases, which includes 6,203 new cases today. That is one confirmed case per 751 inhabitants. As I have explained in another post, the confirmed number of cases vastly understates the number of infections. The number is still going up, despite a nation-wide house arrest.
The original center of the epidemic, Wuhan, had 50,000 confirmed cases in a population of 11 million - that's 4.5 cases per 1,000. However, various studies have concluded that the actual infection rate was several times higher. One study concluded that only 1 in 6 infections showed up in the official statistics; another study estimated that roughly half of the cases were missed. One reason for the under-reporting was that testing was limited to severe cases during parts of the epidemic. This means the actual attack rate was probably between 9 and 30 in a thousand. Without the very strict interventions in Wuhan, it was estimated that the number of cases would have increased 51-fold.

I can perhaps understand if she does not understand the details about epidemic computer models. She is an immunologist, not an epidemiologist or a computer person. Still, it takes about 30 seconds to find out her statement was wrong.

But what really got to me was when she went on how she bragged about
"We do have 19 out of our 50 states that have persistently low level of cases and at this point have less than 200 cases. That's almost 40 percent of the country"
The second part is a bit misleading, to say it politely. Check the by-state list at www.worldometers.info/coronavirus/country/us/, and you'll notice that these 19 states are also the states with the lowest population. Wyoming? Population  579,000 - about 1.4% the number for California. 56 total case, 7 new ones. Alaska? Population 732,000, or about one 25th of the population of New York.  59 cases, 17 new ones today. Other small states show similar growth in the 15% or higher range:
Look at Idaho: the number of cases doubled in a single day! Or just plug the numbers for Vermont (pop. 624,000) into a spread sheet, and extrapolate for 30 days: the number grows to 175,272. That's going from 158 cases to 25% of the state's population in 30 days! That's what exponential growth means!

Relative to the population size, Wyoming has the same number of cases as California. The governor of California clearly understands exponential growth, because he has taken strict measures to stop the spread. It is very sad to see that the US coronavirus coordinator does not understand exponential growth.

Small numbers do not mean that you are safe - they only mean that you are early in the epidemic. The only thing that indicates you are safe is if the numbers do not grow, or if the percentage growth is very low. I am pretty sure Dr. Fauci would be able to explain that to her.

What even worse is that smaller numbers also mean that containment is easier. With 200 cases, it would be possibly to trace all contacts, test them all, and quarantine everyone who tests positive. They did this very successfully in a town of 3000 people in Italy. They knew about just one infected person. They found 89 others the first time they tested, and another 6 the second time. They isolated everyone, and stopped the spread. What do you think would have happened if they had said "that's just one case, no need to worry"? By the way, this little town is an example of an attack rate of 30 in 1,000.

Rather than making statements that can be very easily proven wrong, Dr. Birx time would be better spend to study the actions various government in China, Taiwan, Hong Kong, South Korea, and Singapore have taken to stop the spread of the virus. Those actions differed a lot between countries, and the success varied somewhat, but each of these countries has managed to keep the number of new infections on a manageable level. Despite all the differences, the successful containment measures had a few things in common: they were

  • multi-pronged
  • well coordinated
  • enforced.
In stark contrast, the measures taken in the US 
  • are only based on social distancing
  • differ substantially by state and region
  • often rely on voluntary participation.

Wishful thinking has not worked so far, and will not work in the future. Please, study what has worked and is working in other countries, and implement the measures that work as soon as possible. Only a strong "Hammer" response, followed by a careful "Dance" until vaccines become available in a year or two, will work. Such an approach differs vastly from the current haphazard approach; but it is what will be best not only to minimize deaths and human suffering, but also to bring the economy back on track.