COVID-19 Update; 3/14/2020. A Message From Concerned Physicians

Howard Luks MD ^

[griffin]

Is it time to panic? NO.

This document is trying to help you to understand the situation at hand and not to terrify you. We want to make sure you understand the facts and understand what is at stake. This is a Pearl Harbor moment for our country. We are facing a real threat and we need to face it with all of our resources. When people decry the seriousness of this moment they are steering our country off a cliff, we need everyone to understand that this is important and if we work together to slow the spread we will get through this as we are learning from Singapore, Hong Kong, South Korea, and China.

The effectiveness of our healthcare system to deal with a sudden tsunami of respiratory illness is what is at risk. If our healthcare system buckles under the strain of tens of thousands of patients then we could be looking at a catastrophe.

[LS]

Not sure I understand all this, but my friend Deplorable Greg-—yes, he’s an ObGyn, not a virologist-—has been tracking this and shows that in Iran (and, I think, in Italy) the trend line has begun to curve down.

In the US, IF the China Virus was here in December as most of us think, then the numbers are probably several weeks behind where the real curve is.

[DoodleBob]

I'm not a biostatistician nor do I play on one TV, but if you normalize for population and the stage of the epidemic where each country sits, our active case count will rise very fast, no doubt. However, we are WELL below where other major countries were in their first few weeks of their epidemic. If this trend continues, we'll have the second-lowest per capital case count (the UK is the lowest).

The SURPRISING story should be Switzerland: 3,438 active cases for a (allegedly well-developed) nation of about 8.5MM people. That's the second-highest per-capital case load. It's probably due to a pours border situation with other hot spots Germany, France, and Italy. Go to find the case count by Canton.

[DoodleBob]

I have a cutoff each night (appx 1:30 UTC) when I post my forecast. The numbers jumped after my cutoff.

The error will widen today, no doubt.

[LS]

Interesting. Who woulda thunk it?

[traderrob6]

“I have a cutoff each night (appx 1:30 UTC) when I post my forecast. The numbers jumped after my cutoff.”

It will all come out in the wash so to speak. If the estimates are accurate then will reflect in an underestimate the following day.

[DoodleBob]

Well, the logistic model is dead, Jim. She's undershooting the actual count by 73%.

The Hubei-based forecast is keeping up, but not by much.

I supposed I shouldn't be too hard on myself - it's not every FReeper who forecasts this sort of thing AND publishes their results. Further, it's not in my nature to explain away my own faulty forecast. But today's 76% growth in confirmed cases *MAY* be due to more testing being done here in the US.

The sun will come out tomorrow, and hopefully the rate of confirmed case growth abates. Right now we're looking more like Germany in that regard. I'm a man of my word so I'm going to keep publishing these results until Apr 15; if you'd like to be dropped from this ping please let me know.

Date	Forecast Cases: Hubei Pattern	Forecast Percent Growth	Actual Cases	% diff: Actual-Forecast Cases	Logistic Forecast: 1/(1+exp(-( -3.2392+0.1988*day of the epidemic)))*41500	% diff: Actual-Logistic Forecast Cases
3/14/20	2,999	38.0%	2717	-9%	3,315	-18%
3/15/20	3,552	18.4%	3478	-2%	3,974	-12%
3/16/20	4,376	23.2%	4645	6%	4,748	-2%
3/17/20	6,837	56.3%	6362	-7%	5,650	13%
3/18/20	8,271	21.0%	7769	-6%	6,693	16%
3/19/20	10,202	23.3%	13680	34%	7,885	73%
3/20/20	12,029	17.9%	.	.	9,233	.
3/21/20	13,526	12.4%	.	.	10,739	.
3/22/20	15,264	12.8%	.	.	12,395	.
3/23/20	16,577	8.6%	.	.	14,189	.
3/24/20	18,125	9.3%	.	.	16,099	.
3/25/20	19,408	7.1%	.	.	18,096	.
3/26/20	20,410	5.2%	.	.	20,144	.
3/27/20	20,410	0.0%	.	.	22,204	.
3/28/20	29,488	44.5%	.	.	24,236	.
3/29/20	33,280	12.9%	.	.	26,201	.
3/30/20	34,408	3.4%	.	.	28,066	.
3/31/20	35,590	3.4%	.	.	29,806	.
4/1/20	36,696	3.1%	.	.	31,401	.
4/2/20	37,731	2.8%	.	.	32,842	.
4/3/20	37,945	0.6%	.	.	34,125	.
4/4/20	38,196	0.7%	.	.	35,255	.
4/5/20	38,331	0.4%	.	.	36,238	.
4/6/20	39,201	2.3%	.	.	37,086	.
4/7/20	39,201	0.0%	.	.	37,811	.
4/8/20	39,325	0.3%	.	.	38,427	.
4/9/20	39,630	0.8%	.	.	38,947	.
4/10/20	39,875	0.6%	.	.	39,384	.
4/11/20	40,125	0.6%	.	.	39,749	.
4/12/20	40,320	0.5%	.	.	40,054	.
4/13/20	40,579	0.6%	.	.	40,307	.
4/14/20	40,927	0.9%	.	.	40,517	.
4/15/20	41,047	0.3%	.	.	40,691	.

[aMorePerfectUnion]

I supposed I shouldn't be too hard on myself

You should declare victory and withdraw the troops!

[DoodleBob]

I have not yet BEGUN to fight!

[aMorePerfectUnion]

"I have not yet BEGUN to fight!

"There are many fronts. Choose wisely grasshopper."

[Disestablishmentarian]

What if you tweak the parameters of the exponent in your logistic function to make the curve more closely fit the data that’s coming in?

I don’t know exactly how you came up with those numbers, but obviously they entailed some educated guesswork. Adjusting these factors IN NO WAY invalidates your essential hypothesis—that this SARS coronavirus can be effectively modeled by a logistic function of the same form as other SARS.

The key questions remain:
1) What is the number of actual cases reached by the time the curve flattens?
2) When is this flattening achieved?

[Disestablishmentarian]

PS: Have you corrected your function for there probably being less gin & tonic consumed in America than the U.K.?

[DoodleBob]

[DoodleBob]

The good news is today's US Confirmed Case Count grew by only 41%, which is less than yesterday's massive 76% increase.

The bad news is I still suck as a virus forecaster, but anyone who shoots knows that if you're pulling left, all you need to do is make an adjustment. It seems that the US growth rate pattern since March 14 is more correlated with that of Hubei starting on Jan 24 vs Jan 30 (which is my starting point). However, I'm sticking to my bad forecast for now. Besides, you never know...the Red Sox beat the Yankees after trailing by three games to win the Pennant a few years ago.

I've introduced a new column, of % daily growth in confirmed US Cases. I think this helps flesh out where we were, and where we're going...and how bad I'm failing. Rock on.

Date	Forecast Cases: Hubei Pattern	Forecast Percent Growth	Actual Cases	% Daily Growth in Actual Cases	% diff: Actual-Forecast Cases	Logistic Forecast: 1/(1+exp(-( -3.2392+0.1988*day of the epidemic)))*41500	% diff: Actual-Logistic Forecast Cases
3/14/20	2,999	38.0%	2717	25%	-9%	3,315	-18%
3/15/20	3,552	18.4%	3478	28%	-2%	3,974	-12%
3/16/20	4,376	23.2%	4645	34%	6%	4,748	-2%
3/17/20	6,837	56.3%	6362	37%	-7%	5,650	13%
3/18/20	8,271	21.0%	7769	22%	-6%	6,693	16%
3/19/20	10,202	23.3%	13680	76%	34%	7,885	73%
3/20/20	12,029	17.9%	19285	41%	60%	9,233	109%
3/21/20	13,526	12.4%	.	.	.	10,739	.
3/22/20	15,264	12.8%	.	.	.	12,395	.
3/23/20	16,577	8.6%	.	.	.	14,189	.
3/24/20	18,125	9.3%	.	.	.	16,099	.
3/25/20	19,408	7.1%	.	.	.	18,096	.
3/26/20	20,410	5.2%	.	.	.	20,144	.
3/27/20	20,410	0.0%	.	.	.	22,204	.
3/28/20	29,488	44.5%	.	.	.	24,236	.
3/29/20	33,280	12.9%	.	.	.	26,201	.
3/30/20	34,408	3.4%	.	.	.	28,066	.
3/31/20	35,590	3.4%	.	.	.	29,806	.
4/1/20	36,696	3.1%	.	.	.	31,401	.
4/2/20	37,731	2.8%	.	.	.	32,842	.
4/3/20	37,945	0.6%	.	.	.	34,125	.
4/4/20	38,196	0.7%	.	.	.	35,255	.
4/5/20	38,331	0.4%	.	.	.	36,238	.
4/6/20	39,201	2.3%	.	.	.	37,086	.
4/7/20	39,201	0.0%	.	.	.	37,811	.
4/8/20	39,325	0.3%	.	.	.	38,427	.
4/9/20	39,630	0.8%	.	.	.	38,947	.
4/10/20	39,875	0.6%	.	.	.	39,384	.
4/11/20	40,125	0.6%	.	.	.	39,749	.
4/12/20	40,320	0.5%	.	.	.	40,054	.
4/13/20	40,579	0.6%	.	.	.	40,307	.
4/14/20	40,927	0.9%	.	.	.	40,517	.
4/15/20	41,047	0.3%	.	.	.	40,691	.

[Robert DeLong]

I think your forecast is good. Lots of unknowns, and initial testing is going to throw it off as it was starting from behind the curve. So deviations are to be expected.

[Disestablishmentarian]

I am fascinated by your analysis, even if the last 2 days are not fitting well. So I set up my own Excel doc with your equation.

QUESTIONS:

1) How did you arrive at the asymptotic value of 41,500? I thought we were solving for the max cases ... but now I realize that assumption came from a priori reasoning.

2) Obviously the epidemic hit the U.S. earlier than your current assumption, which places patient “Zero” around March 10th ... whereas there were at least 70 known cased by about March 2nd or so. So I’m thinking March 14 should be more like DAY 15 in the growth equation.

3) Can you give me some insight where the other 2 parameters in the exponent came from?

4) I get a better fit to the data, especially today’s totals, using March 14th as day 15, setting the MAX CASES=500,000; resetting your .1988 factor to .2754311; and resetting your 3.2392 factor to 9.0.

Even so, this assumes there was a degree of under-testing until the last 2 days.

NOTES:
500,000 confirmed cases (if that were the number) would certainly not overwhelm the health system, and would probably result in just few thousand deaths, allowing POTUS to declare victory.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
It is conceivable the Trillion-Dollar BAILOUT CHECKS ALONE could cost about $1 billion per death.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Someday we will truly understand the growth function of this virus, and be able to assess the magnitude of the (over)reaction versus what we could have legitimately faced with normal safeguards like washing hands and quarantining actual cases.

[DoodleBob]

Thank you. I'm honored.

1+2) I got 41500 by extrapolating the Hubei confirmed case growth rates from 1/30 on the US confirmed case counts circa 3/10. Clearly, that was an error. In retrospect, I picked the wrong Hubei starting date; the recent fit is much better if I indexed the growth rate starting from 1/24 in Hubei. If I reset the growth rats like I said in #1, the US case count forecast grows to about 181k.

3)I eyeballed the growth rates to build a data set of high and low growth flags by date, and modeled a logistic function. There are several 'free' ones online. Since I modeled growth rates vs the actual counts I didn't have to correct for autocorrelation in the data.

I can, of course, 'fix' the model but I'm a man of my word. Once this thing becomes way bad I can reset the forecast. Besides, I've already posted my forecast so it's all transparent.

[DoodleBob]

Despite my forecasts continuing to be way off, the rate of growth in the US slowed for the second day in a row. It's also worth noting that NY accounts for half of the growth in the US over the past three days. If I removed NY, the US Confirmed Case Count is 11,468, which is within spitting distance of my Hubei-based forecast AND the logistic model.

Date	Forecast Cases: Hubei Pattern	Forecast Percent Growth	Actual Cases	% Daily Growth in Actual Cases	% diff: Actual-Forecast Cases	Logistic Forecast: 1/(1+exp(-( -3.2392+0.1988*day of the epidemic)))*41500	% diff: Actual-Logistic Forecast Cases
3/14/20	2,999	38.0%	2717	25%	-9%	3,315	-18%
3/15/20	3,552	18.4%	3478	28%	-2%	3,974	-12%
3/16/20	4,376	23.2%	4645	34%	6%	4,748	-2%
3/17/20	6,837	56.3%	6362	37%	-7%	5,650	13%
3/18/20	8,271	21.0%	7769	22%	-6%	6,693	16%
3/19/20	10,202	23.3%	13680	76%	34%	7,885	73%
3/20/20	12,029	17.9%	19285	41%	60%	9,233	109%
3/21/20	13,526	12.4%	24493	27%	81%	10,739	128%
3/22/20	15,264	12.8%	.	.	.	12,395	.
3/23/20	16,577	8.6%	.	.	.	14,189	.
3/24/20	18,125	9.3%	.	.	.	16,099	.
3/25/20	19,408	7.1%	.	.	.	18,096	.
3/26/20	20,410	5.2%	.	.	.	20,144	.
3/27/20	20,410	0.0%	.	.	.	22,204	.
3/28/20	29,488	44.5%	.	.	.	24,236	.
3/29/20	33,280	12.9%	.	.	.	26,201	.
3/30/20	34,408	3.4%	.	.	.	28,066	.
3/31/20	35,590	3.4%	.	.	.	29,806	.
4/1/20	36,696	3.1%	.	.	.	31,401	.
4/2/20	37,731	2.8%	.	.	.	32,842	.
4/3/20	37,945	0.6%	.	.	.	34,125	.
4/4/20	38,196	0.7%	.	.	.	35,255	.
4/5/20	38,331	0.4%	.	.	.	36,238	.
4/6/20	39,201	2.3%	.	.	.	37,086	.
4/7/20	39,201	0.0%	.	.	.	37,811	.
4/8/20	39,325	0.3%	.	.	.	38,427	.
4/9/20	39,630	0.8%	.	.	.	38,947	.
4/10/20	39,875	0.6%	.	.	.	39,384	.
4/11/20	40,125	0.6%	.	.	.	39,749	.
4/12/20	40,320	0.5%	.	.	.	40,054	.
4/13/20	40,579	0.6%	.	.	.	40,307	.
4/14/20	40,927	0.9%	.	.	.	40,517	.
4/15/20	41,047	0.3%	.	.	.	40,691	.

[DoodleBob]

Despite my forecasts continuing to be way off, the rate of growth in the US slowed for the second day in a row. It's also worth noting that NY accounts for half of the growth in the US over the past three days. If I removed NY, the US Confirmed Case Count is 11,468, which is within spitting distance of my Hubei-based forecast AND the logistic model.

Date	Forecast Cases: Hubei Pattern	Forecast Percent Growth	Actual Cases	% Daily Growth in Actual Cases	% diff: Actual-Forecast Cases	Logistic Forecast: 1/(1+exp(-( -3.2392+0.1988*day of the epidemic)))*41500	% diff: Actual-Logistic Forecast Cases
3/14/20	2,999	38.0%	2717	25%	-9%	3,315	-18%
3/15/20	3,552	18.4%	3478	28%	-2%	3,974	-12%
3/16/20	4,376	23.2%	4645	34%	6%	4,748	-2%
3/17/20	6,837	56.3%	6362	37%	-7%	5,650	13%
3/18/20	8,271	21.0%	7769	22%	-6%	6,693	16%
3/19/20	10,202	23.3%	13680	76%	34%	7,885	73%
3/20/20	12,029	17.9%	19285	41%	60%	9,233	109%
3/21/20	13,526	12.4%	24493	27%	81%	10,739	128%
3/22/20	15,264	12.8%	.	.	.	12,395	.
3/23/20	16,577	8.6%	.	.	.	14,189	.
3/24/20	18,125	9.3%	.	.	.	16,099	.
3/25/20	19,408	7.1%	.	.	.	18,096	.
3/26/20	20,410	5.2%	.	.	.	20,144	.
3/27/20	20,410	0.0%	.	.	.	22,204	.
3/28/20	29,488	44.5%	.	.	.	24,236	.
3/29/20	33,280	12.9%	.	.	.	26,201	.
3/30/20	34,408	3.4%	.	.	.	28,066	.
3/31/20	35,590	3.4%	.	.	.	29,806	.
4/1/20	36,696	3.1%	.	.	.	31,401	.
4/2/20	37,731	2.8%	.	.	.	32,842	.
4/3/20	37,945	0.6%	.	.	.	34,125	.
4/4/20	38,196	0.7%	.	.	.	35,255	.
4/5/20	38,331	0.4%	.	.	.	36,238	.
4/6/20	39,201	2.3%	.	.	.	37,086	.
4/7/20	39,201	0.0%	.	.	.	37,811	.
4/8/20	39,325	0.3%	.	.	.	38,427	.
4/9/20	39,630	0.8%	.	.	.	38,947	.
4/10/20	39,875	0.6%	.	.	.	39,384	.
4/11/20	40,125	0.6%	.	.	.	39,749	.
4/12/20	40,320	0.5%	.	.	.	40,054	.
4/13/20	40,579	0.6%	.	.	.	40,307	.
4/14/20	40,927	0.9%	.	.	.	40,517	.
4/15/20	41,047	0.3%	.	.	.	40,691	.

[gas_dr]

I think NY has reached unsaturated testing. In other words, zero order kinetics was at play at the number of cases was limited by the number of tests. in the case of NY, there is now more tests than patient demand. As such there is probably increased cases that are not touching the healthcare system, which explains the mortality rate of 0.5%. This is in line with what is prognosticated and severe flu (yes I will be called a flu bro by the FEARPERS).

There is a Growth Facto <1.0 for the first time today worldwide.
It appears using todays numbers that the curve is flattening
Additionally there is not as many cases as originally deared
NY is a hot spot, clearly, but I think superior testing shows true rates.

Worldodometer was a little hinky today — multiple changes. I think there will be some variance tomorrow, but I think there is significant likelihood that the curve is probably flattening

Lets hope

[DoughtyOne]

Thanks for the updated status.