Thread by @CJinKnoxTN

The Real Challenge for NY State: Hospital Beds (not Ventilators). Also, the dog that did not bark ;)

Another thread about COVID-19 and hospital beds in NY and using data to track a new expert model. (Note: previous thread was on NY ICU beds and ventilators). /1

Tuesday (3/31) the White House revealed they were following a model from Univ of Washington (UW): covid19.healthdata.org/projections This new model answered the question about ventilators, but like the model I used in yesterday's thread it showed other breathtaking projections. /2

With regard to NY ventilator needs, the UW model definitively contradicts Cuomo’s dramatic claim that he needs 40,000 ventilators. It projects 9,666. Cuomo says he already has 10,000. At peak in next week or two, it shows that NY State will need apx 74K hosp beds!!! /3

Interesting side note: My model (hybrid from HGHI) assumed an ICU adm needed a ventilator. The UW model separately projected ventilators at about 80% of ICU admissions. But back to the gut punch: 74K beds!! Do you know what that means? /4

That means the UW model effectively projects that by Apr 9, there will be 526,000 positive cases of COVID-19 (73,629 HospAdms/0.14 AdmRate) in NY over a 2 wk period starting this **past Thursday**. It effectively means they will test 1.3M (526K/0.40 PosRate) ppl in NY. /5

Stop. Drink that thought in. They are going to test 1.3M ppl in NY in 9 days?!?! But are they only going to test in NY? What about everywhere else? Absurd. The daily confirmed case increase right now is just under 13%. It must double to 25% & hold that to get to 500K by 4/9. /6

To be fair, the White House said it may be sometime in the next 2 wks. So, it will have to jump tomorrow & hold at 16% daily growth for 2 wks. It compounds, so any significant miss is great news. Reminder: you count only the last 12 days since the avg hosp stay is 12 days. /7

If I'm wrong, and the UW model proves to be accurate. Then the real problem in NY is hospital beds, not ventilators. Cuomo challenged hospitals to free up capacity, & that is captured in my hybrid HGHI model as "Potential Hospital Beds Available". Let's assume they get there. /8

Also, Cuomo got 4K beds from FEMA and 1K more from US Navy Ship Comfort. He's asked for 4K more beds in for more temp hospitals. Let's assume he gets those. /9
thehill.com/homenews/state…

If he gets those, there's still a deficit of apx 34,000 beds. If the apex is indeed 7-14 days away, and he expects 73,000 cases (check that, he's still claiming 140,000!) why on earth has he not begged for 34,000 or 100,000 more beds?!?! His ask for 4K more is the tell. /10

It's the dog that did not bark. Neither he nor the White House truly believe he's going to need more than 40,000 beds. In fact, they wouldn't be cutting it that close. They don't think they will need more than 30,000. /11

Could you imagine the press conferences after the fact if Cuomo only asked for 4K more beds when his own projections were for 140K beds and he only had 36K available? Or Trump being asked why he didn't build more when he knew they were short 34K? /12

Not a chance. I'm not a prophet, nor the son of one, but I think the peak hospitalization will be closer to 27K (250K positive cases at peak). In any case, I will be praying every day for my fellow man and esp NYC, & I will be updating my model with any data I can scrounge. /13

That was one of the other big takeaways from today's WH briefing. The UW model has a real-data feedback loop. They update with data twice per day. That means the only surviving assumptions over time are correct ones. That's a welcome development. /14

Seriously - My thoughts and prayers are with NY and other affected areas. We are all from NY this next week. /15 END

PS - My prediction is not driven by the hybrid model. I've been tinkering with it and scrounging for data in hopes it might reveal something significant. It was my gut 2am and how there is no clamor/uproar asking the most obvious ques about 34K bed shortage. /16 END