The doubling period of Covid-19 is between 1.8 and 4.1 days where large scale quarantine measures aren't being taken [1][2]. The conference is in 43 days so without large scale quarantines and stopping large events then the number of cases in the US could grow to as many as 800,000 assuming a doubling period of 3 days. It seems therefore fairly obvious that most large events in April will have to be cancelled.

[1] https://link.springer.com/article/10.1007%2Fs15010-020-01401... [2] https://onlinelibrary.wiley.com/doi/abs/10.1002/jmv.25723

Thanks for sharing the journal articles.

According to [1], on January 31, 2020 US had 6 cases of Covid-19. And according to several news sites, as of March 2, 2020, there are 101 cases.

If I did the math correctly, that seems to be doubling about every four days or so, which lines up with the doubling period mentioned.

Its difficult to tell because the US hasn't been doing as much testing as other countries. You can look at the curves for various countries at https://covid.bio/ Most countries that haven't initiated large scale quarantines have a doubling period of between 1.5 and 5 days

On the other hand, one wonders if that curve measures not infection rates, but testing kit manufacturing output.

I think the numbers can't be looked at that way for a variety of reasons, primarily the fact that we are rapidly expanding testing capability in the US right now, so we will see a rapid increase in the number of diagnosed cases. However that does not mean there is the same rapid increase in the number of cases so quickly. Basically we are going to be catching up with cases from the past 2-4 weeks in the coming week.

Your March 2 numbers include Diamond Princess cases (45?) which have not been in contact with the rest of the population.

No there are other estimates of a few hundred cases in the bay area that are sensible. Or at least they were a week ago. To get two in-the-wild cases serious enough for hospitalization and testing (overriding the CDC testing criteria at the time) you'd expect there to be at least around 100-200 cases at the time of hospitalization.

the us has tested less than 500 cases in total

If we have about 100 cases today, and that doubles every 4 days, we'd have about 60k cases in 37 days.

    (6 + 31) / 4 = 9.25
    2 ^ 9.25 = 608
    608 * 100 ~= 60k

When I wrote "6 + 31 days" I meant "6 (initial) cases, 31 days later". I thought that was clear in context.

Ah, yes `6 * 2 ^ (31 / 4)` is much larger than 100. Good point.

> Currently, there have not been any COVID-19 cases in Pennsylvania

No boundaries to stupidity. Such arguments makes no sense at all, better check how many people have been tested in PA. While everyone cancelling large scale events there are few mohicans going forward seeing no risk.

Maybe buy a few temperature guns and point them at people's foreheads before they enter?

Due to the long incubation period, infected people can be infectious even while not presenting symptoms or relatively asymptomatic.

Bai, Yan, Lingsheng Yao, Tao Wei, Fei Tian, Dong-Yan Jin, Lijuan Chen, and Meiyun Wang. “Presumed Asymptomatic Carrier Transmission of COVID-19.” JAMA, February 21, 2020. https://doi.org/10.1001/jama.2020.2565.

Guan, Wei-jie, Zheng-yi Ni, Yu Hu, Wen-hua Liang, Chun-quan Ou, Jian-xing He, Lei Liu, et al. “Clinical Characteristics of Coronavirus Disease 2019 in China.” New England Journal of Medicine 0, no. 0 (February 28, 2020): null. https://doi.org/10.1056/NEJMoa2002032.

"Patients often presented without fever, and many did not have abnormal radiologic findings. ... Fever was present in 43.8% of the patients on admission but developed in 88.7% during hospitalization."

While that's certainly true, viral load is important and this likely still would be an effective risk reduction measure.

The biggest news here is that pycon still uses blogspot.

Eagerly looking for the updates on this!