Latest COVID-19 Projection from IHME

The media has pretty much buried the fact that recent US Covid-19 hospitalizations have dropped by 70-75% since the 18th of April. There's no way they can handle having it look like things are improving in at least one measurable, better to look at overall totals or point out spots where things aren't getting better.  Here's a link to the weekly hospitalization chart...  gis.cdc.gov/grasp/COVIDNet/COVID19_5.html (I couldn't figure out how to display the chart).  It'll be interesting to see if the hospitalizations are on an uptick when the next week or two numbers come in.

Take a look at the disclaimer on your site. They only track about a dozen states, about 10% of the population of the US.

it's a fair question. I don't need defender, not that I don't appreciate it. 

I was probably too condescending in my reply too. calling it "poop" data wasn't fair. I was probably in a bad mood when I posted that. I apologize. 

I would give you a link, but I am literally not allowed to. It is a proprietary tracking system that has no link to give. Yes, I am well aware that sounds suspiciously like "my uncle works for Nintendo..." however, there is reporting done out of the system, and it shows up every day on the OHA daily reports, found here: govstatus.egov.com/OR-OHA-COVID-19. you can use this link to see they are off on their testing projections. they are not that off on deaths though... sadly. 

Where I was unfair is not being more descriptive as to what I mean. My umbrage was pegging resource capacity at 210 ICU beds. That number is an estimation, and the site makes it clear. It is fine a site estimates this, as bed capacity is notoriously difficult to get. But the issue is I don't think it is really a great thing to track in the manner they are tracking it. Hell, we are having a very difficult time doing it at OHA. 

This is because in reality, a bed is a bed is a bed in a hospital. The state does not license bed types. it licenses beds. period. There is no separate license for an ICU bed. What makes an ICU bed different than a regular bed is the level of staffing it has and the equipment that supports it. These things can, and do, fluctuate. a hospital can scale ICU beds up and down to the maximum they have staff and equipment for.

The site is trying to estimate it, but the estimate is probably off. We have an excess of 700 unused ventilators in the state. Importantly, something not reported publicly or on this site, is we have a large availability of negative flow rooms. rooms in which the pressure in the room is lower than the outside pressure, such that when you open the door the outside air sucks in. This is pretty critical for isolating contagious individuals. There is capacity to expand ICU beds as needed, above and beyond the 210 bed availability. 

I think it is important to track patients needing ICU care, but I do not believe it is an appropriate measure of resource capacity. It is too flexible. Total beds available, in general, is a better measure. Because once you run out of ANY beds, you are gonna start having problems. The site does do this, and I have no problems with that. If you read a hospital has 20 ICU beds and 19 are occupied, it isn't necessarily cause for alarm. That hospital can just convert some beds over with a semi-trivial amount of effort in most cases. However, a hospital that has 200 total beds and 195 are occupied, it might be time to panic. 

Hospital notoriously run both lean, and under report their actual capacity. This is for many reasons. 1. it isn't sustainable to run at max. 2. it isn't safe. 3. it costs a poop load to staff heavy.

ICU care requires more nurses per patient, it isn't ideal to have a high percentage of your beds be ICU care. you have to staff at a higher ratio. 

Also I would remiss if I didn't give Johns Hopkins a shout out. I have worked with researchers from there on a couple projects and they are all great. They are also free. free help is fantastic, because contractors and consultants cost a s%#tload. So there you go, Johns Hopkins is out there saving Oregon taxpayer money!

Honestly, that doesn't look that ominous. It is weird how the goal posts have changed. When the pandemic was declared the idea was to flatten the curve to make sure hospitals don't get over run and to give time for treatments and vaccines to be developed.  Those curves all look real flat, no logarithmic increases to be seen.  Now it seems the goal posts are set not at flattening the curve, but in zero deaths (we don't require zero deaths from any other preventable disease but this one...can you imagine if in the 90's we said we were going to prevent all HIV deaths by stopping people from having any sex for 2 years?  Would have worked, I guess, maybe a little harder to get people to go along then wearing masks ;^). Anyway, death sucks, I get that, but millions have died in preventable car accidents, but we don't spend trillions on death proof cars and roadways (which is very possible, if you just spend trillions on them).