In which jon-nyc stakes out an unconventional opinion on the Covid-19 outbreak
-
wrote on 8 Apr 2020, 03:22 last edited by jon-nyc 4 Aug 2020, 22:07
Despite all the models saying otherwise, I don't think this will peak in April.
I'm not even confident New York will peak in April.
When I say 'we won't peak in April' what do I mean? I mean the number of active cases (total cases - recoveries) won't peak in April (even if they dip for a time). And by implication the utilization of healthcare resources won't either.
The models disagree with me. Cuomo has now, two days in a row, expressed cautious optimism that we (NYS) are nearing a plateau.
I think they're wrong. (Obviously I hope they're not.)
Why do I think they're wrong? I don't think our collective social distancing measures are enough to get R0 under 1.
Why do I think that? Two reasons.
(1) Wuhan
(2) Europe
1 - Wuhan did two phases of social distancing / lockdown. The first phase was arguable stricter than the one we are on now.
In their first phase, they blocked people from leaving Wuhan, closed public transit and vehicle traffic, made mask wearing compulsory, canceled public events, and had confirmed or suspected cases self-quarantine.
Note we are only doing the last two.
According to a study by Chinese and US researchers, that first phase brought R0 down only to 1.26.
They then implemented the second phase which included a full quarantine of confirmed or suspected cases (took them out of their homes to a quarantine site), started monitoring all residents’ temperatures, and enforced very strict rules about who could leave the house and when (IIRC, one person per household could leave every other day for food).
That same study says the second phase got their R0 down to about 0.35.
2 - No European country has yet hit a peak of active cases. Not even Italy, which did a national lockdown about 2 weeks or so before we really did. Like us, Italy and Spain have brought down their daily case increase, but they're still increasing faster than old cases are resolving. Even if they do peak, if their R0 is not <1, they will simply continue their increase from a lower base.
Again, this is a heterodox opinion. I am probably wrong, and I certainly hope I am wrong. Nevertheless this is where I think we really are.
-
wrote on 8 Apr 2020, 03:28 last edited by
Jon, my hubby has been talking about this, too. He also thinks there's much longer until the peak for many places in the US, including NYS.
-
wrote on 8 Apr 2020, 03:51 last edited by
A recent Imperial College work estimating the effective R value in European countries. In all countries except Norway, they estimate the value to be more likely above 1 than below. Though the 50% confidence interval contains 1 for all countries except Sweden.
Peruse the graphs on pp 7+
Also some scary estimates on actual infection rates on p 6, which are more than a week old!
-
wrote on 8 Apr 2020, 03:52 last edited by
Jon you make a good argument. I wouldn’t bet against your position.
For the dumb people in the room (me) who aren’t following the lexicon perfectly, is “peak” defined as total active cases or when growth factor (cases today / cases yesterday) is consistently <1.0?
-
wrote on 8 Apr 2020, 03:58 last edited by
Look at my third sentence!
What I'm talking about peaking is active cases. So the number of people who have the disease at a particular time.
Something I didn't mention, though I've stated in another thread, is that the peak active cases occurs after the peak in new cases, since cases last a couple of weeks. If that's not intuitive to someone I'm happy to graph it out.
-
wrote on 8 Apr 2020, 11:24 last edited by Loki 4 Aug 2020, 11:31
I will only say that the IMHE model stays very consistent and in fact overall US death projections are down again this morning.
Vents in NY are of course projected to be 20% of Coumo’s main pitch lat week.
Hospital resources across the country, which are generally their visits, hospitalizations and revenues are down over 60%. It’s a bloodbath.
I think there is a big lag in test data which you are not seeing.
For IHME to forecast peaks this week in deaths and hospital resource this week and you to say more than two weeks from now, that is a huge spread.
-
wrote on 8 Apr 2020, 11:29 last edited by
I have nothing to add, other than this was the first thread I opened this morning.
Thanks for starting my day off on a cheerful note, Jon.
-
wrote on 8 Apr 2020, 11:33 last edited by jon-nyc 4 Aug 2020, 22:08
Davis - Like I said, I’m hopefully wrong.
I’m violating one of my own rules rules here. Experts in a topic may well be wrong, but usually not in a way that is obvious to the layman.
But for now I will stand by this.
-
Davis - Like I said, I’m hopefully wrong.
I’m violating one of my own rules rules here. Experts in a topic may well be wrong, but usually not in a way that is obvious to the layman.
But for now I will stand by this.
wrote on 8 Apr 2020, 11:37 last edited byIf IMHE model doesn’t stand up to the next two or three days you won’t have to look at any more and all those people will be a distant memory. You are basically calling them soon to be irrelevant. It’s too close for the event itself for them to recover. I’ve seen lots of hubris in my life, I’m not betting in these personalities being so blind to the data. All models are wrong, only data speaks, here the data should be screaming at them.
-
wrote on 8 Apr 2020, 12:12 last edited by
No, if the models don’t hold up they’ll learn from that and adjust them. Try looking at this through a lens of science instead of politics. Turn the TV off if necessary.
-
No, if the models don’t hold up they’ll learn from that and adjust them. Try looking at this through a lens of science instead of politics. Turn the TV off if necessary.
wrote on 8 Apr 2020, 12:28 last edited by@jon-nyc said in In which jon-nyc stakes out an unconventional opinion on the Covid-19 outbreak:
No, if the models don’t hold up they’ll learn from that and adjust them. Try looking at this through a lens of science instead of politics. Turn the TV off if necessary.
Jon, try considering this. IMHE uses actual deaths as the basis of the model and not cases which are way underreported.
But this is your thread and you have set the goal posts and are apparently calling the shots at the skeptics. Should be fun.
-
wrote on 8 Apr 2020, 12:36 last edited by
I’m not ‘calling the shots’ I’m thinking out loud.
Why is all of this a ‘who’s up, who's down’ question for you? Don’t you have any native curiosity in the underlying topic?
-
wrote on 8 Apr 2020, 12:38 last edited by
The testing lag is a good point.
But other countries‘ data should also lag, no?
Taiwan and SK had about a 10-14 day gap from peak new cases to peak active cases, testing lags and all. Why wouldn’t we see that in our data?
-
I’m not ‘calling the shots’ I’m thinking out loud.
Why is all of this a ‘who’s up, who's down’ question for you? Don’t you have any native curiosity in the underlying topic?
wrote on 8 Apr 2020, 12:39 last edited by Loki 4 Aug 2020, 12:41@jon-nyc said in In which jon-nyc stakes out an unconventional opinion on the Covid-19 outbreak:
I’m not ‘calling the shots’ I’m thinking out loud.
Why is all of this a ‘who’s up, who's down’ question for you? Don’t you have any native curiosity in the underlying topic?
Ok
-
The testing lag is a good point.
But other countries‘ data should also lag, no?
Taiwan and SK had about a 10-14 day gap from peak new cases to peak active cases, testing lags and all. Why wouldn’t we see that in our data?
wrote on 8 Apr 2020, 12:48 last edited byJon. Are you looking at the R(0) data by country, trended.
Also have you looked at it by state?
I can show you where to find that if it will help.
-
wrote on 8 Apr 2020, 12:51 last edited by
Please do!
-
wrote on 8 Apr 2020, 12:52 last edited by
@jon-nyc said in In which jon-nyc stakes out an unconventional opinion on the Covid-19 outbreak:
Please do!
https://91-divoc.com/pages/covid-visualization/
The second chart.
-
wrote on 8 Apr 2020, 12:58 last edited by
No R0 estimates there.
-
wrote on 8 Apr 2020, 12:59 last edited by
The trend is still up pretty much everywhere.
-
wrote on 8 Apr 2020, 13:02 last edited by
@jon-nyc said in In which jon-nyc stakes out an unconventional opinion on the Covid-19 outbreak:
No R0 estimates there.
Sorry, I thought you were looking for cases by state, not R0.
