Wednesday, April 1, 2020

COVID-19 Diary

March 31. Starting a little late here - but might as well jump in. Almost three weeks ago, I stopped going in to my office. The office officially closed on Monday, March 16 - but I  figured it was safer to stay at home ASAP, so I stopped commuting on Wednesday, March 11.

Connor stopped being able to attend school on March 13 (which I think is the first day Ed worked from home) and Helen's school system closed on the following Monday, March 16.

On March 13, our governor announced that schools would be closed through the remainder of the school year, and the students and teachers would instead engage in "distance learning". Hearing that schools were closed for the duration of the year was hard.

Yesterday, March 30, Virginia got a stay home order, which doesn't change our lives too terribly much since we've been mostly hunkered down - but the stay home order is in place until JUNE 10. JUNE 10 is a LONG ways away. That was hard to hear.

April 1. There is an interesting model that attempts to project hospital needs, based on a few inputs including when schools were closed, when a stay home order was put in place, when non-essesntial services were shut down, and when travael was severely restricted. Also feeding into the model are reported cases and death. The analysis is state-by-state, and in some cases, it's pretty comforting.

But not today.

Yesterday, the data did not include the date of Virginia's stay-home order, which should have the effect of pushing our peak resource need out considerably. It also wasn't totally up to date on deaths and confirmed cases.

The model suggested that peak usage would be May 28 in Virginia. And, because I have family in Kansas, I recorded the peak usage day in Kansas as well - April 27. The US peak usage day was projected to be April 15 - but given the unevenness of this mess, it's hard to derive much meaning from that date.

Overnight, the model was updated. For the US, the peak usage date was pushed back one day - to April 16. But here in Virginia, our peak usage day was MOVED FORWARD! Our peak usage predicted date is now May 20. Kansas was pushed back a day to April 28.

So...when an input that should have pushed our date back was added - it wasn't enough to outweigh the inputs that move the date forward, which is presumably identified cases. Maybe the first date was false, because testing has been so inconsistent, so the change doesn't really mean anything. But when I look at the numbers - and I really see it as there being very little else we could do, personally - today slots solidly into the "hard days" column. These days are thankfully rare, but really overwhelming.







No comments:

Post a Comment