March 24, 2020

Today’s numbers. In Westchester, 18,000 tests have been administered (3282 yesterday), and we have 3891 positive cases (remember there’s a lag between tests administered and results). Increase of about 1000 since yesterday. NYC now has 14,904 cases, NY State has 25,665 total. Cuomo says numbers are doubling every three days, so they think the apex will be sooner but much higher than previously thought — with a need for between 110,000 and 140,000 hospital beds (the state has 50,000) in the next two weeks. From the raw numbers, it looks to me like Westchester cases may be doubling every two days versus three, but either way, even three more doublings means 30,000+ cases in Westchester in the next week to 10 days, which is sobering if there’s a 10+% hospitalization rate. But again, we’ve been in lockdown a bit longer so hopefully we’ll see our numbers flatten before too long (the problem is that the parameters for getting tested keep shifting so even if we see the numbers appear to flatten who knows it we are comparing apples to apples)? If you have not watched Cuomo’s daily news conferences, you should.

Comments [Names abbreviated to protect the innocent]

AC Blocking you for 30 days.

TF similar to retail store closings…

PB Thank you lance!

AO Take a look at this article: https://www.nytimes.com/2020/03/20/health/coronavirus-data-logarithm-chart.html

Lance Koonce So I switched to a log scale for each chart, and added the US as well, below. What does not make great sense to me is that my US chart ends up looking a lot flatter than the one in the NYT article, even if I just use the data through March 18 like they did. It looks to me like the reason is that the Excel chart has wider tick marks for the date axis, which makes the line flatter. Any data experts here who can explain whether there is any “right” way to format the date axis here?

SS Alternatively, a bar graph of new cases by day will give a sense of when things turn. When new cases start to decline, that will be an indication of the curve flattening. Theoretically, that bar chart will look like a bell curve when all is said and done.

SB SS assuming testing access remains constant…

PB They are also turning away people from getting tested because of lack of tests

SC He said we need 30,000 ventilators within 3 weeks.

VS Let visualizations speak for themselves.

RG Maybe some optimism from the positive growth rates falling across the board?

RG Something weird for 3/18 though, NYC had more new cases than NYS that day?

SN Lance excellent work. Let me point out that the death statistics from Westchester in NYT are definitely inaccurate. I know of at least 4 deaths at [REDACTED] the hospital I work at