Growth rate for new cases nearly static for 4 days.

[Ms. Toad]

Still growing exponentially (daily multiplier >1)
Still growing at a significantly slower rate of increase than it was 18 days ago
But the daily multiplier been mostly static (going up & down within a .95% range (less than 1% variation))

3/22 - 1.385797497
3/23 - 1.303732189
3/24 - 1.253526924
3/25 - 1.244204075
3/26 - 1.252510592
3/27 - 1.218774507
3/28 - 1.186812131
3/29 - 1.149452168
3/30 - 1.15344921
3/31 - 1.150667708
4/1 - 1.140417971
4/2 - 1.138946898
4/3 - 1.131837616
4/4 - 1.123379552
4/5 - 1.081308594
4/6 - 1.090090384
4/7 - 1.090821898
4/8 - 1.08640492

Dr. Acton (Ohio) mentioned a national change change in diagnostic criteria - which would identify possible and probable cases based on clinical criteria. I don't know that these will be included in the cases or not. If it is included in our diagnoses, I'd expect to see a dramatic, short-duration increase - like the 2/12 jump in China when it changed testing criteria. I haven't found any broader mention of this change, or of how these new cases will be treated for reporting.

I really love our team in Ohio - apparently a lot of others do, as well. Two have fan clubs:

Dr. Acton (Head of department of health & former Obama volunteer*) https://www.facebook.com/groups/500716313939425
Marla Berkowitz (ASL interpreter extraordinaire & married to a woman*) https://www.facebook.com/groups/163190034745240/

*Added because of the many DU threads trashing red states/red governors. We don't always have to be moral enemies.

Feel free to drop in at 2:00 PM daily, 6 days a week and meet these two wonderful leaders! Even when the news is grim, it is a bright spot in the day, in large part because of these two women.

[Demovictory9]

1. good news

[Ms. Toad]

2. I wish it was still declining -

but it has declined significantly, and at least is not increasing as several suggested would quickly happen when the growth rate started to slow ~ March 22.

[ffr]

3. We still don't have enough testing to know who's infected. It's a disgrace!

Before I believe in a possible pause, I'd like to have a better understanding of the population that isn't being tested. I'll bet the numbers are going to spike unexpectedly again, if not new cases, sudden spikes in symptomatic deaths from patients who haven't been diagnosed with Covid-19 and will therefore not be counted as Covid-19 deaths.

We are not out of the woods by any means.

[Ms. Toad]

5. You appear to be misunderstanding what I have posted.

This is showing a decline in how quickly the number of cases is growing. It is NOT showing "out of the woods, or a lower number of new cases each day. It is showing that we are no longer pedal to the metal heading toward a brick wall. We still have our foot on the accelerator - and until the daily multiplier is less than one, we will still get more new cases tomorrow than the number of new cases we had today we're still climbing the hill that is getting steeper (more cases each day) - we're just not climbing the mountain any more. The curve we're on now suggests we'll still have more cases each day than the prior one until around the 15th or 16th.

[uponit7771]

6. Would a sub par per capita testing rate of symptomatic effect the multiplier any? Thx in advance

[Ms. Toad]

8. Impossible to be absolutely certain without a lot more number crunching.

If the testing is growing at a slower exponential rate than the actual new cases it theoretically could - but the curve is very smooth. I'd expect case identification curve that was capped by the number of tests available to be herky-jerky. We'd grow rapidly, then bump up against a testing cap, more tests would become available and we'd jump and then grow smoothly until we hit the testing cap, be stuck there a while, new tests become availble so we'd jump again.

The other thing I would expect to see is that if the growth in new cases was being capped by testing limitations, we would be seeing hospital and ICU admissions closer to the original models but the number of confirmed cases would be disproportionately low compared to the hospitalizations. Our hospitals would be running out of beds and vents at the rate that was predicted before we started social distancing - and they aren't. In high density places the situation is still horrendous. And Ohio is expected to be pretty bad over the next week (our curenlty projected peak).

But I'm not hearing, for example, the horrendous choices they had to start making on a daily basis in Italy. If the new case numbers were being capped by testing I would expect everyone to be smiling at how well we've flattened the curve - BUT - Italy-like situations in our hospitals. I'm not seeing that.

[uponit7771]

9. Could an infinite supply of test affect the smoothness of the curve also? My understanding is

Last edited Thu Apr 9, 2020, 09:52 AM - Edit history (1)

... there are enough test relative to the 5 million Pence said would be available on March 16th, its just they're not being given because Trump doesn't want the numbers to go up.

So there are "jumps" in Texas and Missouri for instance but not from availability of test but from the criteria the CDC and state level has given for testing.

This is one of the reason I"m thinking the "curve" reporting should ... NOT ... be nationally centered or focused because like electoral college elections ... national numbers mean VERY little while swing state or in the case of CV19 .. hot spot ... numbers mean everything.

Trump is right in saying that everywhere is not flared by that's like someone with CV19 saying they're in the clear its just their lungs are feeling tight.

[Ms. Toad]

13. The challenge with local focus is that the data set

is more wobbly because of the smaller numbers (just like the national set is until we start to hit the thousands of cases).

I looked at a curve for a jurisdiction with fewer than 300 cases and the daily multiplier was all over the place. It looked more like Trump's signature than a nice smooth curve.

So, except for states with tons of cases, I wouldn't trust my curve to predict a peak. The folks who get paid to do this have access to a lot more data than I do, so they can probably pull it together for a more reliable model.

I don't really think it gets them anywhere to try to hold the numbrs down. COVID 19 deaths aren' really invisible, from the front line stories I'm hearing. So if they are trying to keep the positives down, it will be pretty cleare the numbers are inaccurate based on the deaths that can't be hidden that easily (especially since the vast majority of deaths occur in the hospital, where tests are being done - so a higher proportion of them will be tested)

In terms of opening things back up, it will be almost impossible to do responsibly unless there is adequate testing to identify hot spots - on a local basis (after the local peak has passed). Ohio is trying to figure tha out, and has acknowledged that without testing, opening back up will be a much slower process.

[uponit7771]

18. I agree, Analyst are not trying to hold numbers down but Trump is ... NO DOUBT ... holding numbers

... down cause he's an idiot and truly believe he'll choke America for to death for dollar.

Looks like NY is going to test part of the 100 - 300 dead people a day they're finding in homes so if a good number of those are CV19 we'll see a change if

I do agree on the testing, hell is going to break lose in the market if Fauci says May opening depends on testing.

[Ms. Toad]

20. Fauci doesn't dictate what local states do,

that is in the purview of the State departments of health (in most cases). So what he says may not matter much in terms of the reality of what is opened.

[ProfessorGAC]

22. It Doesn't Here In Illinois

Pritzker & Nzeki have no reason to fudge the numbers. Even in your state with an R governor, I don't see them withholding the true numbers. DeWine has been very aggressive in trying to tamp down the spread.

[Ms. Toad]

25. Agreed.

No one on the Ohio team (with the exception of AG Yost - who seems completely disconnected from all of this aside from his push to close planned parenthood/abortion providers) is treating this as a political game. All of them are insistent that our actions be driven by science - part of which is knowing real numbers about how much infection there is - and where. We don't have enough tests, we need more tests, we're working as hard as we can on being able to do more tests is a daily drumbeat in the press conferences.

[uponit7771]

35. 100% ... didn't think about this, Red Don can say all clear but it'll be local areas that give the

... real all clear.

Politically Trump will say it's all the democrats fault for the economy crashing but economically that doesn't get us out of the soup.

Yep, then Red State govs will give all clear and reinfect everyone ... jus damn ...

I think they want Wisconsin for Nov 3rd ... jus damn

[Midnightwalk]

10. Maybe a dumb question.

I’m not sure how to ask this right.

When you say there will be more cases until around the 15th or 16th how does that vary by state?

I think the most cases are in New York, New Jersey and Michigan? Does that mean the data more accurately reflects what’s happening there or can I be more hopeful for the rest of the country?

I’ve seen some web sites that show different states peaking later, but I don’t have the background to tell if they are accurate.

[Ms. Toad]

11. The curve I'm working with is the national curve

The more data, the more likely it is that it will be reasonably accurate. But you are correct that on a state-by-state basis states will reach their peak on different days - based on when the virus was seeded there, whether/when they implemented social distancing rules, whether they tend to have a lot of international travelers, how dense their population centers, etc.

I don't know how accurate - their modeling is significantlty more sophisticated then mine. I'm just plotting the data I have on a graph and fitting a curve to it. It has been pretty accurate for teh nest 1-3 days, and generally accurate for predicting big events, like the peak - for example my national peak has consistently matched the predicted national peak predicted by most models.

[SheltieLover]

15. They should, imo, be not only testing more

But also dxing AND documenting on clinical sxs.

Lots of false negative tests & viral load must have developed to a certain level to show a + result.

[ananda]

16. Agree.

There are way more infections and deaths
than the numbers show, and we all know
why

[duforsure]

17. Their preventing mass testing to hide the real numbers

trump already has said he's not responsible for getting them tests, PPE supplies, ventilators, or aid. He's done everything to not help anyone except for a few who have kissed his ass. trumps response to this virus has been he's been completely MIA. Now its no testing just go back to work daily propaganda promotions , next will come threats to Governors by withholding funds if they don't force people back while its still going on. trumps no response to this shows he could care less about the American people's health . We already know he cares nothing about the truth.

[greenjar_01]

30. Texas test per million is a disgrace; Ohio is on the lower-end tests per million*, but not too bad

California is also low on tests per million.

*for bigger states

I suppose we'll find out eventually whether tests per million is a relevant measure. It seems that wherever we have > 5,000 tests per million, we see bigger numbers, but that varies a bit too, and will probably even out as states start getting the full effects. I don't see how <5,000 tests per million is giving us any real workable data, much less <4,000. Texas and Georgia are real eyebrow raisers in this regard.

[krawhitham]

4. Yeah, but we are not testing much. Ohio ranks 40th for tests per capita

[Ms. Toad]

7. This isn't Ohio data -

it is national data - so for purposes of interpreting the data, the adequacy of Ohio's testing is not relevant.

That said, Ohio is testing as much as it can, given the shortages of testing supplies/ingredients. For some reason being a red state governor did not mean a windfall of testing supplies.

No question but what there should be tons more testing. But the trend follows the same trends I saw in China and Italy. The # of death is trending the same way - just a week or so later. Since I'm not even suggesting this is the entire number of deaths, and the number of tests is increasing while the number of new cases is not growing as quickly as it was when our testing was even more limted, I'm confident it is a real trend and not a measure of being capped by the number of tests we are doing.

[FreeState]

12. Compared to national testing numbers

Interesting when compared to national testing numbers.

https://covidtracking.com/data/us-daily

3/22 1.25814880573515
3/23 1.24021709224201
3/24 1.23350735133627
3/25 1.22278175510251
3/26 1.23202508943568
3/27 1.20659955558808
3/28 1.17405881911741
3/29 1.13000744864783
3/30 1.13652837369535
3/31 1.11010166650086
4/1 1.09640913580529
4/2 1.10238381364967
4/3 1.10184516165673
4/4 1.16285431202223
4/5 1.08476329257234
4/6 1.08304332724945
4/7 1.07664551234386
4/8 1.06964339947358

[Ms. Toad]

14. I'll have to play with that.

Thanks for the link.

[Squinch]

32. Great info. It took a week to double. Much better than the 3 days it was

taking in the beginning.

A fraction of the actual total cases, but it shows a good trend.

[Yavin4]

19. The first step in proper data analysis is proper data collection.

Without comprehensive national testing, we don't know the real numbers. We don't anything truly. We only know when people show up really sick at the hospital, and the problem with that is that people can be asymptomatic and still carry the virus.

My fear now is that limited data collection which has lead to limited analysis will be used to push us all back to normal too quickly.

[Ms. Toad]

21. Absolutely - but there is nothing I can do about that

And using the data that is collected, even in its flawed state has reasonably corellated with real world observations in three countries now, I see no reason to stop looking at it and using it for the purposes I've been using it for - to get a pulse on what is going on. But just as it would be malpractice to diagnose coranary artery disease based on taking someone's pulse, it would be governing malpractice to open up the country.

My fear is that, just as people underestimated what exponential growth looks like in the first place, they will do the same when the cases start to grow exponentially once restrictins are eased.

FWIW, the people with actual medical, scientific, or mathematical knowledge have linked opening things back up to adequate testing.

[Yavin4]

23. We're dealing with a president that only believes in narratives not saving lives.

He will use limited data to push his narrative.

[Ms. Toad]

27. Fortunately, since local restrictions on movement are local decisions

his narrative can't actually open the country up. So the real risk is in those states whose governors refuse to exercise independent thought (and adjacent states their population is likely to travel to frequently).

[ElementaryPenguin]

24. We hardly test anyone, and therefore have NO IDEA how many Americans are infected...

How many are asymptomatic carriers, who needs to be isolated, etc.

Trump's federal government's response could not possibly be any more incompetent and negligent!!!

[Squinch]

28. BUT, is it safe to think that the fraction tested is relatively stable?

If so, these numbers, though wildly rough estimates, still show a slowdown.

And that means deaths should slow in a couple of weeks.

Problem is that some states will take this slowdown as a signal to open up again and that will restart the clock for those areas.

[DBoon]

26. When the first derivative is zero

and the second derivative is negative, you have reached a downward inflection

Did I recall college calculus correctly?

[Ms. Toad]

29. Yes. n/t

[gibraltar72]

31. Look at the number of tests

Pitiful.

[Ms. Toad]

34. Agreed. n/t

[NoMoreRepugs]

33. I'm having a difficult time wrapping my head around the numbers.

Rural areas in red states coupled with rabid red state governors who worship tRUMP do not provide me with much confidence that the reported statistics are anything close to accurate.

[Ms. Toad]

36. In terms of absolute numbers of infected - the lack of testing means the data

isn't much good to give us a good feel for how many cases there actually are.

But based on other things we can observe (like how crowded our hospitals are, whether the number of deaths has a similar curve that lags the number of cases curve by 1-2 weeks, I'm relatively confident that even that inadequte data identifies trends.

[NoMoreRepugs]

41. Appreciate your insight. Once there is a better handle on all the deaths both at home and in

hospitals NOT being tested for Covid19 I'm of the opinion both the mortality rate and mortality figures are going to be staggering.

[Ms. Toad]

42. Agreed - the toll will be staggering.

[captain queeg]

37. This is good but we really don't know for sure what the numbers are

With an administration actively trying to hide and subvert the data.

[Ms. Toad]

38. Absolute numbers, no.

But I'm confident that we can determine trends from what we're seeing, based on confirmation things we can observe directly.

[Takket]

39. have to think the number stagnating te last 4 days......

is new cases slowing in states that shutdown two weeks ago while states that didn't shut down are accelerating, balancing each other out.

[Ms. Toad]

40. That's possible -

When he new case growth rate all over was so stong, the decline in more populated areas that were shut down would outweigh any increase in less dense populations. And just like the beginning of the exponential curve in growth - the squashing benefit in the populated areas is less pronounced near the end of the squash - while the places not under lockdown may be picking up speed at a higher daily multiplication rate and - as you suggest - may be balancing it out.