Reply With Quote
New York is actually doing a phenomenal job. .8 percent mortality rate vs. Italy's near 10 percent.
Its tough to find science guys and tell em forget about the science.
But orange guy will find a gadfly.
New York is actually doing a phenomenal job. .8 percent mortality rate vs. Italy's near 10 percent.
Yep, and the rate isn't growing. Case number is because testing is growing.
Updated numbers:
March 23: 15.08%
March 24: 15.02%
The woman, Dr. Brix is serviceable. She's a Never Trumper, but, is mature & keen enough to "just let it go" & drink from the silver cup as long as she can & they'll let her.
In other words why isn't Houston currently Italy? So without even considering possible temperature interactions, Because the time of introduction of a sufficient number of infections to promote fast community spread wasn't until later. In the early days of exponential spread, minor events have a huge impact. For me in chemistry, it was mostly luck. The doubling time of particle growth will be a distribution. In other words, you have an average value but this is affected by random events in both directions.
I assume (this is not close to my work even though there are math parallels) that in epidemiology this happens in much the same way. For instance, lets say you get two visitors from Wuhan, and one goes to Seattle and one goes to Boston. The one from Seattle doesn't have much personal contact with anyone else, and just goes about their business They manage to spread the infection but at a slow rate. Well the that branch of the tree is growing at the regular rate. But the dude who goes to Boston goes to a convention and actually has close contact with a lot of people. Oh and this dude is the kind of personalty that likes to get really close when they talk to people or whatever. Well, his branch EXPLODES.
Early on, this is a big event, right? One place grows way faster than another based on randomness. But overall, the math is the same and because its exponential growth even if you have far less numbers, its still more important to look at doubling times and whether or not there's community spread. So my main point is this: You cannot take sollace that there are limited numbers in Houston (for example) because there's enough for it to be spreading in the community there which means that its going to spread and its going to do so fairly rapidly. Even with the measures we have in place now. They aren't very rigid yet, so its still going to spread and the numbers will grow quite dramatically.
So back to the weather. its already been rather warm in SE Texas, Lousiana, and Florida. All of these places have community spread and it doesn't seem to be very slow. Its growing at an exponential rate in every single one of these locations. The flue is already in decline in these locations due to the weather but COVID19 is doing the exact opposite and it's doing it incredibly fast. Yes, the numbers are low, but relatively low numbers mean little compared to the exponential rate that its growing at.
Its not about the current numbers, its about the growth rate.
i hope mid and co are right tbh... if we can get out of this economic stall relatively quickly, better chance of a recovery without seeing massive unemployment as a result. unemployment hurts everybody (less consumer purchasing power means providers of goods/services scale back as well)
Testing percentages are a really bad way to extrapolate the growth curve which is why no one is using that. What epidemiologist have you seen look at testing percentages to make a point about the rate of growth? In a limited testing environment there is enormous selection bias built in. For instance here in Brazos county, the only people getting tested as of yesterday were those who were symptomatic, and traveled or had contact with someone who had tested positive. I assume its pretty similar in Houston.
Now if you did random testing among the population, then that metric would have more value. There have been calls to do this in places and I believe California actually did this (and the numbers they got were why Newsom reacted with lockdowns). But just looking at percentages when you're selecting a subset of the population doesn't tell you anything about the growth in the population at large.
One way you can get a better handle on the total rate for a population is through working back based on the number of hospitalizations. If you expected X% of infections to result in a hospitalization at Y time, then you can look at backwards and figure it out. Its pretty simple algebra at that point, but you're making rate assumptions so the math is fuzzy. I saw a guy who did this for NYC this past weekend and I'll see if I can dig up the medium post.
looking at testing for that purpose only makes sense if the people tested are a random, representative sample of the larger population, which isn't the case. you cant be extrapolating like that
But you should know that different environments can promote or re growth rate. Community spread will happen much faster in environments favorable to it. Houston's population density is pretty low vs. other American cities with more than 500K people.
https://en.wikipedia.org/wiki/List_o...lation_density
Then there's the mass transit variable (how much does Houston use it? Do they even have a system)? The weather variable (which I feel you think is still totally inconclusive, but I think it does matter), social behavior variable, etc, etc.
California's policy is to only test people with obvious symptoms. They're even advising people who might have symptoms, but aren't severe (we still have no clue how many people who have this only have very mild symptoms) enough to warrant a hospital visit to stay home. So basically, California is just testing obvious cases. I'm sure many flu or severe cold sufferers are getting tested with negative results.
So my point here is that even with this targeted testing, the majority of people with Covid like symptoms don't have the virus. But let's just hope those with mild symptoms stay put.
Well yeah the case number is growing because of more testing but why on earth do you think its not growing exponentially? Nothing in the data agrees with you. Put the data on a log plot and its clear as day that its damn close to a straight line. If it wasn't, even with more testing coming online, it wouldn't show that growth.
This is just 100% wrong.
And yet, they still have exponential growth which is the issue. I guess this concept really is just hard for people to grasp.
Does anyone here who's not really worried about all these possible hotspots wonder why every epidemiologist who's out there is terrified of the growth rates of this virus? Like honestly, why are the all freaked out over this if something like the heat is just going to kill it? Or why are they asking for even stronger measures than what is occurring right now? Why do the people most familiar with viral spread seem to have the greatest sense of fear of this thing?
As a climate scientist, I 100% feel for them because I'm used to people who aren't experts thinking big problems are NBD.
Agreed 100%. I think what you're saying is that a one-size-fits-all approach does not work. You have to look at the facts and cir stances, to borrow a phrase lawyers throw around. And, as you rightly point out, those facts and cir stances include the amount of personal exposure had between the two patient zeros you laid out.
But those facts and cir stances include, well, other facts and cir stances. People like me and mid have been arguing for a more holistic assessment of each state, city, community, etc... That approach would take into account factors such as population density, climate, public transit, etc... The point being that you cannot look at what's happening in Italy and say "just wait until that happens in X or Y location."
By way of pure anecdote, they shut down the Houston rodeo -- an event with hundreds of thousands of people exposed to each other -- because they found out that someone who had attended the cookoff -- an event with approximately 74,000 people out at the football stadium's parking lot -- had coronavirus (https://www.click2houston.com/rodeo/...r-coronavirus/). This happened in late February, but you haven't seen the exponential growth here that you'd normally expect.
Here too I think there are a number of factors at play. Yes, the Texas numbers are rising -- but they're far out of sync with what you see in Washington, New Jersey, and New York. Those numbers are, at least in part, attributable to the rise in testing. But, I think the overall point is that while you will still see growth - the rate at which that growth occurs will be markedly slower in places like Texas and Florida because of the facts and cir stances of those locations. What mid and I are speaking to is the rate in which growth occurs. And that rate seems to be slower in hotter and less dense locations as compared to the Pacific Northwest or NYC.
I don't see how 10 percent increases (in case rate) each day is "exponential growth?" Here's Italy's plot:
https://imgur.com/a/pyOUcOn
Last two days, their growth rate has been 8 percent over the previous number. Exponential growth is a constantly doubling, 1, 2, 4, 8, 16, 32, 64, etc. Or are you defining exponential growth differently? Or are you plotting it on a longer timeline? And their growth rate day over day has declined over the past week.
If it's what I'm thinking of, Mid posted it and Medium pulled it down ...
the reality is New York curve is making Italys look like a flacid penis tbqh
they gotta get wrecked
https://coronavirus.1point3acres.com/en
You can switch the plots to a log. They aren't perfect straight lines, but you can fit an exponential function to those plots with a very high level of variance accounted for. That is a textbook exponential function. Anyone saying its not exponential growth is 100% wrong.
But an exponential doubling over a 3 month period won't be a doomsday scenario, i.e. South Korea. They're averaging a 100 new cases since they flattened, so that means they'll get case doubling from now in 90 days. Maybe this is where we're talking past each other. What timeline are you plotting for this exponential growth? Do you expect/fear consistent case and death doubling every week, month?
They're not far out of sync though. I think this is what you're missing. You see numbers that are far lower and you think that it means there's way slower growth. I see that and I think it means that there was a later introduction. This is not to say that every place is going to be a worst case scenario. We've already put measures into place to hopefully avoid that, but honestly those measures aren't even showing up in teh data we currently see so its hard to know if we've bent the curve enough or not.
Without a doubt, actions like the rodeo action have likely helped. The question is have they helped enough? Without much of a buffer, its really hard to say. It will in some places, it will not in other places, but the fact that there are relatively large numbers prior to lock downs (and the data we currently see is reflective of the situation weeks ago, not now) all over the place does not bode well.
South Korea flattened the curve by combination of massive testing and extreme great tracing(immediate notification/tracking/quarantining of infections. everyone would get notified).
USA has neither so there wont be any similarities to the flattening of the curves
Oh 100% that the goal is to slow the doubling time. But nothing we've done yet is reflected in the data for most places across the US since the lag time for the data is weeks. Plus, we're not doing what SK did. Not even close, so its really not realistic to expect results on the same magnitude they have achieved.
yup you cant compare us with SK because we are not doing anything coose to what they did
apples and oranges tbqh
There are currently 12 users browsing this thread. (0 members and 12 guests)
Posting Permissions