Thoughts on Ohio's COVID-19 numbers: Are we flattening-the-curve, or is something else at play?
Easter Sunday, April 12, 2020: I have been thinking about the Ohio COVID-19 numbers, and how they just don't seem to be making sense. I posted my thoughts to my Facebook page, but I feel others may wish to consider my thoughts as well, so I'm posting them here, on my website. If you wish to discuss, feel free to contact me.
-Tom Lusch <TomLusch@me.com>
-------------------------------------------------------------------------
[Posted to Facebook on Wednesday morning, April 8, 2020, at 7:26 a.m. EDT]
Awakened early by thunder & hail, I got to thinking more about Tucker Carlson’s show last night (April 7th: Tucker: Experts can't predict when coronavirus pandemic will end: A close look at the data suggests the peak of the epidemic in New York may have already passed). One can find that segment here…
https://video.foxnews.com/v/6147891073001
In that show, Tucker questions the predicted numbers, and the decisions that were predicated upon those predictions.
That got me thinking. I decided to do my own “back of the envelope” calculations, looking back at Ohio thus far.
I’ll begin with what the Columbus Dispatch reported that Governor DeWine & Dr Acton shared with us nearly one month ago on Thursday, March 12th. See…
“A fifth Ohioan tested positive for COVID-19…cases are expected to double within six days…confirmed numbers are just a small fraction of the individuals who are infected already…Acton said 1% of Ohio’s population, more than 100,000 people, were carrying the virus as of Thursday…”
Interestingly, as I check the coronavirus.ohio.gov website dashboard, it indicates an official number of cases as being 357 on March 12th, as opposed to the 5 cases they stated that day. I’m figuring that is probably due to delays in test kits not having been analyzed as of when they spoke. I’ll use that official 357 number of cases and double it every six days, and see how it compares to the official numbers thus far. As a point of reference, I recall reporters advising that cases were doubling every 3 days in hot spots, so doubling every 6 days certainly seems to be on the conservative side.
Thursday 3/12/20
Unofficial infection estimate = 100,000
Official Ohio cases = 357
Official Ohio deaths = 0
Wednesday 3/18/20 (1st doubling)
Unofficial infection estimate = 200,000 (100,000x2)
Estimated cases = 714 (357x2)
Official Ohio cases = 1,265
Official deaths = 2
Tuesday 3/24/20 (2nd doubling)
Unofficial infection estimate = 400,000 (100,000x2x2)
Estimated cases = 1,428 (357x2x2)
Official Ohio cases = 2,375
Official Ohio deaths = 18
Monday 3/30/20 (3rd doubling)
Unofficial infection estimate= 800,000 (100,000x2x2x2)
Estimated cases = 2,856 (357x2x2x2)
Official Ohio cases = 3,708
Official Ohio deaths = 74
Sunday 4/5/20 (4th doubling)
Unofficial infection estimate = 1,600,000 (100,000x2x2x2x2)
Estimated cases = 5,712 (357x2x2x2x2)
Official Ohio cases = 4,723
Official Ohio deaths = 153
Saturday 4/11/20 (5th doubling)
Unofficial infection estimate = 3,200,000 (100,000x2x2x2x2x2)
Estimated cases = 11,424 (357x2x2x2x2x2)
Official Ohio cases = TBD
Official Ohio deaths = TBD
It certainly appears that we Ohioans are definitely “flattening the curve.”
But what do these numbers say about our Case Fatality Rate (CFR)? CFR is the proportion of deaths from the disease compared to the total number of people diagnosed with the disease.
As of yesterday, April 7th, the COVID-19 Dashboard indicates an official 164 “Total Deaths,” and an official “Total Cases” of 4,782.
If one were to calculate a CFR using those numbers (number of deaths, divided by the number of cases), one would come up with a CFR of 0.03429527, or 3.4%. That sounds really high. And being that the number of deaths lag the number of cases (i.e. a percentage of people recently diagnosed will likely die), the CFR would likely increase.
However, we are all well aware that due to lack of testing kits and/or the reagents needed for those test kits, that testing for COVID-19 has only been done on a small minority of people who are infected.
Speaking of CFR, what is the rate for the flu that we regularly deal with? Taking a look at the CDC website at…
https://www.cdc.gov/flu/about/burden/index.html
…refer to “Figure 1: Estimated Range of Annual Burden of Flu in the U.S. since 2010.” There we find this statement…
“CDC estimates that influenza has resulted in between 9 million – 45 million illnesses, between 140,000 – 810,000 hospitalizations and between 12,000 – 61,000 deaths annually since 2010.”
If we take the low end numbers (12,000 deaths divided by 9,000,000 cases), it results in a CFR of 0.13%. Similarly, using the high end numbers (61,000 deaths, divided by 45,000,000 cases), that similarly results in a CFR of 0.13%.
But being as we know the number of official COVID-19 cases in Ohio is way low due to the CDC bungling the test kit rollout, let’s look at our Ohio CFR from a different angle.
Let’s assume that Dr Acton’s remarks on March 12th, as reported by the Columbus Dispatch reporter, is a correct estimate (“…more than 100,000 people, were carrying the virus as of Thursday…”).
Also, to be really conservative, let’s assume that no other Ohioans (magically) have gotten infected since March 12th, when that unofficial estimate was that 100,000 Ohioans were infected. Within that near month, the disease would’ve likely run its course for the 100,000 infected.
As of yesterday (4/7/20), we’ve had 164 Total Deaths due to COVID-19.
Therefore, 164 deaths divided by 100,000 infected.
That comes to 0.00164.
That would indicate a CFR of 0.16%.
That CFR is pretty darn close to the 0.13% “CDC Estimated Range of Annual burden of Flu in the U.S. since 2010.”
What do you think?
“It's tough to make predictions, especially about the future.”
-Yogi Berra
-------------------------------------------------------------------------
[Posted Friday afternoon, April 10, 2020, a little after 2:00 p.m. EDT]
Sorry it has taken me so long to get back to you Paul. I agree that the true CFR won't be known for quite some time. Nevertheless, the concern about the CFR is obviously the driver of our nationwide shutdown.
I am absolutely with you on us owing to our fellow citizen the physical distancing that we've been doing. Nevertheless, questioning what our leaders actions are predicated upon, and being skeptical about those predictions, is exactly what I wish to see in a journalist. That is why Tucker's show on 4/8 rang so true.
As I hear about the decreasing _predicted_ number of deaths, I have not been able to help but wonder about the predictions. Nationwide I've heard the predicted number of deaths due to COVID-19 ranged from 1.6 to 2.2 million initially, and then dropped down to estimates of 100K to 200K, and not long after that estimate dropped down to 60K. Those are incredible adjustments, especially since I thought, at least on the last two estimates, that their models had taken into account the "social distancing" our nation has accomplished with our shutdown.
Looking at Ohio, just about a month ago (March 13th) it was written in The Columbus Dispatch "A USA TODAY analysis of American Hospital Association bed data shows that if U.S. rates of critical illness are the same as in China and Italy, there could be six patients for every hospital bed in the country. The same analysis finds that in Ohio, there could be more than 14 patients for every available hospital bed." [1]
With estimates from the experts of 14 patients for every available hospital bed, it would be lunacy for our governor to have not dropped everything he was doing and focus entirely on shutting down Ohio to stop the spread. To his credit, Governor DeWine proactively embarked on just that, and that was before anyone was determined to have died of the coronavirus here in Ohio. I have admired his courage and his proactiveness when given such grim estimates by his director of health and the modeling experts.
On March 16th it was written, "At least 2% of Ohioans who get the virus and become ill with COVID-19, about 2,000 people, could die before the pandemic ends."[2]
Then on April 2nd the estimated number of deaths doubled. It was written, "The leader in the study of infectious diseases at Ohio State University believes social distancing is having a positive impact but still predicts there could be up to 4,200 deaths during a surge expected to start later this month across the state.”[3]
On April 4th (just 6 days ago), we see written, "Modeling the state is relying upon anticipates that the peak of the virus will bring up to 10,000 new cases a day...The peak would occur in mid- or late April."[4]
But come April 5th, we're beginning to get news more like this...
"...the public might understandably be confused by estimates that swing widely, with ranges of possible cases and deaths differing by thousands or even millions."[5]
"The research team plans to publish a white paper this week to explain in more detail the ranges and the methods used to predict them."[5]
This week the numbers appear to be leveling off here in Ohio. From coronavirus.ohio.gov I presently see...
Date: Total Cases Total Deaths
04/05/20: 5,314 175
04/06/20: 5,518 191
04/07/20: 5,686 202
04/08/20: 5,812 219
04/09/20: 5,876 230
04/10/20: 5,878 231
The numbers (thankfully) certainly aren't jibing with the predictions here in the Buckeye State.
Now I'm going to turn on the TV and catch their Friday 2 o'clock update.
[1] https://www.dispatch.com/news/20200313/hospitals-prepare-for-coronavirus-but-may-need-more-beds-as-disease-peaks
[2] https://www.dispatch.com/news/20200316/coronavirus-in-ohio-math-behind-infections-spread
[3] https://www.dispatch.com/news/20200402/coronavirus-in-ohio-dewine-extends-stay-at-home-order-until-may-1-as-deaths-increase-25
[4] https://www.dispatch.com/news/20200404/coronavirus-in-ohio-cases-top-3700-deaths-pass-100
[5] https://www.dispatch.com/news/20200405/predicting-coronavirus-cases-deaths-is-tricky-but-models-improving-experts-say
-----------------------------------------------------------------------
[Posted during the wee hours of Easter Sunday morning, April 12, 2020]
I woke up early on Wednesday morning, questioning the predicted number of COVID-19 cases for the State of Ohio, and was wondering what the Case Fatality Rate (CFR) might be, based upon estimated and current numbers.
I did some back-of-the-envelope type calculations based on infections doubling every six days, and found we have been impressively flattening-the-curve here in the ole Buckeye State.
You can view my simple calculations of those doubling estimates on my 4/8 post. Here are my simple estimates to complete the 4/11 data...
Saturday 4/11/20 (5th doubling)
Unofficial infection estimate = 3,200,000 (100,000x2x2x2x2x2)
Estimated cases = 22,848 (357x2x2x2x2x2x2)
Official Ohio cases = 6,250
Official Ohio deaths = 247
As a reminder, my "Unofficial infection estimate" is based upon an estimation of the number of Ohioans infected, as suggested by Dr Acton one month ago today, wherein she said it was likely that 1% of Ohio's population, or more than 100,000 people, had already been unknowingly infected. Based upon that, by this 5th doubling period, without mitigation, we would've likely had 3.2 million of our 11,747,694 population[1], or just over a quarter of us Ohioans (27%), infected. (Note that with doubling and without social distancing measures, 100% of Ohioans would be projected to be infected by April 23rd!)
I also doubled, every six days, the official number of cases that were known as of 3/12. With that, we see that estimated number is 3.6 times greater than the number of official cases shown as I view the dashboard just now (22,848 vs 6,250).
Both ways of looking at these numbers seem to indicate that Ohio's STAY-AT-HOME mitigation measures are working. In fact, they seem to be working far better than anyone expected.
Then I happened across a newscast that called into question whether Ohio's strict social distancing measures are what really flattened the curve[2]. The One American News Network asserted that our governor said that our strict social distancing measures reduced the cases of new infections from a projected unmitigated 62,000/day, to a mere 1,600/day.
So I decided to look into OANN's assertions. This network questioned whether the timing of the STAY-AT-HOME order made sense with regard to these lower numbers. They showed a chart, but it was blurry. I found the chart on the presser of 4/8[3]. Sure enough, it clearly indicates we knocked the numbers down from approximately 62,000/day unmitigated new cases, to a projected approximate 1,600/day new cases.
But as I look at the cases per day on the Ohio dashboard at this moment, <https://coronavirus.ohio.gov/wps/portal/gov/covid-19/dashboards> I see that from April 10 to April 11, we added ONLY ONE CASE! (Total Case Count went from 6,249 to 6,250).
See picture at...
https://06927bd2-a-62cb3a1a-s-sites.googlegroups.com/site/thomasglusch/file-cabinet/Ohio%20COVID%20Total%20Cases%20comparison%20202004120148.jpg?attachauth=ANoY7cp06YyAka4HsS2CLbPFT23-941yQXjk_YNeYwVAQwD7cenfNG0XmYSICln4LWULCRlVIAgbgoDukLzdrQ_3lI5iKQFwpzDmGC1FwFBrK3wb4AInObpa4XpzLpe0ufcMkEx1Ty0Z3pCqaJWi-Nb4bd82AbALfsvgtujtAzE5VpLRAns25yDhrRs-s5_R7l8mLgoMdMzVc8c5M5ljJ1KOkQJGfkyPjIszePKqur3QBEIlQkQ2--7t-jaihyGXlv5r8GO-luaLfm2bKduF8ynm-qZoCVG84w%3D%3D&attredirects=0
That's incredibly amazing!
OANN asserted that the unmitigated peak of a projected new cases/day load of 62,000 had been expected to occur on March 22nd. That correlates when I look at the clean copy from the presser. They also state that the STAY-AT-HOME order[4] didn't go into effect until March 23rd (a day _after_ the unmitigated peak was projected). Sure enough, that is what it clearly shows on page 12 of the presser!
The one thing that OANN didn't mention is the various orders that occurred prior to the statewide STAY-AT-HOME order. For instance, a little over one week prior to the blanket STAY-AT-HOME order, Governor DeWine declared a STATE OF EMERGENCY for Ohio (3/14). Then Dr Acton immediately limited mass gatherings of 100 or more people. That was shortly followed by limited nursing home visits, required screening for admission to Psychiatric Hospitals and Youth Services Facilities, followed by limited access to jails and limiting liquor type sales. Then they stopped the election from taking place on 3/17 as they tightened the screws on mass gatherings (defining such as 50 or more). They also halted elective surgeries, and limited access to nursing homes, and then closed down business like barber shops. But sure enough, the really big hammer of a STAY-AT-HOME order didn't take effect until 3/23, the day _after_ the massive peak of projected cases/day.
I must agree that it sounds pretty strange to take credit for dropping the infection rate so dramatically based upon their orders, especially when the REALLY BIG HAMMER didn't go into effect until a day _after_ the projected peak of 62,000 new infections per day.
I'm thinking something else is causing the numbers to be so incredibly different than the models have predicted.
[1] https://worldpopulationreview.com/states/ohio-population/
[2] https://youtu.be/H10xw28AMyQ?t=48
[3] https://coronavirus.ohio.gov/static/slides/04-08-20-Presser-Final.pdf
[4] https://coronavirus.ohio.gov/static/DirectorsOrderStayAtHome.pdf
-----------------------------------------------------------------------
[Posted Tuesday, April 14, 2020 @ 5:55 p.m. EDT]
During the wee hours on Sunday morning, I wrote, “I’m thinking something else is causing the numbers to be incredibly different than the models have predicted.”
I just learned of a British Medical Journal article entitled “Covid-19: four fifths of cases are asymptomatic, China figures indicate.” [1]
The author writes how new evidence emerging from China indicates that the large majority of coronavirus infections do not result in symptoms. Apparently this China data began appearing on April 1st. An epidemiologist wrote ‘…this suggests the virus is everywhere. If—and I stress, if—the results are representative, then we have to ask, ‘What the hell are we locking down for?’”
Part of my personal experience is what has had me wondering if the coronavirus has already ripped thru Ohio (as well as through our entire country), without our infectious disease experts being aware of it.
It was January 29th that my wife came down with a bad cold that took her a couple weeks to shake. We didn’t think much about it at the time. Then, a couple of days after a Friday wherein I had taken my 13-year old son and his friends out for sled riding (school had been called off due to snow), he came down with a cold (Sunday February 9th). It began as a cough for him, but he also had a very runny nose. We kept him home from school the following Monday & Tuesday. At the time I correlated the cold as being associated with him being outside for several hours, playing in the snow. Then, by the next Sunday (Feb 16th), I came down with a cold. For me it was a really odd cold, nothing like I ever recall experiencing before, as I didn’t have a runny nose or the usual stuffiness in my head. It certainly wasn’t the typical “head cold.” It was simply a cough. It hung on for nearly five days. None of us experienced a fever in conjunction with our colds in early February.
Being as I had picked up on the coronavirus story in late January, and followed it closely all along, when I experienced my cold, and especially since it had began as a cough, I immediately wondered if I could’ve picked up the dreaded coronavirus. Of course my wife rolled her eyes at me as I wondered aloud about that, as, at the time, there had been a mere 15 cases reported in our entire country, and none in Ohio (the first Ohio case wasn’t reported until March 10th). Thus, I agreed at the time…it simply didn’t seem plausible. But it sure had been a weird cold for me.
However, it was at that March 12th news conference wherein Dr Amy Acton, our Ohio Director of Health, stated that she estimated that 1% of Ohio’s population (more than 100,000 Ohioans), were likely already carrying the virus. That rekindled my wondering if possibly I had picked up the SARS-CoV-2 virus a month prior.
I’ve also talked with friends in which their entire family experienced rather difficult colds during early February.
Therefore, based on Dr Acton’s remarks, the British Medical Journal article, and my family and some friend’s personal experience, I cannot help but wonder if COVID-19 has already infected the majority of us here in Ohio…and that such may be why the predictive models have been so incredibly off.
Watching the news the other night, I learned that the National Institute of Health is just now beginning a study to ascertain if there are already many of us who are now immune.[2] I’ve sent off an email to see if I can participate.
[1] https://www.bmj.com/content/369/bmj.m1375
[2] https://www.nih.gov/news-events/news-releases/nih-begins-study-quantify-undetected-cases-coronavirus-infection
-----------------------------------------------------------------------
[Posted Sunday, April 19, 2020 @ 1:30 a.m. EDT]
As I reflect on the number of coronavirus cases in our local news over the past few days, I can't help but be left wondering how effective Ohio's actions to guard against this SARS-CoV-2 virus have been thus far.
One of the earliest actions to protect our vulnerable population from this COVID-19 disease in Ohio began on March 11th, when Dr Acton immediately limited access to Nursing Homes across our state. Nevertheless, around five weeks later in The Columbus Dispatch (Thu, Apr 16th) they wrote, "More than 100 nursing homes and assisted-living facilities in Ohio report that 826 residents are affected by the novel coronavirus. That includes reports from eight central Ohio facilities, and the total makes up about 10% of all cases in the state." [1]
Wow...10% of the number of cases in our state have occurred in nursing homes. That's incredible, as nursing homes account for such a tiny percentage of our state's population.
Also, consider Ohio prisons. On March 12th in The Columbus Dispatch it was written, "With inmates in many state prisons tightly packed, all visitation by family members and friends at adult and juvenile detention facilities has been suspended, as have activities led by outside volunteers." [2]
Yet, with those restrictions put into place a little over five weeks ago, on April 18th we see the headline, "Coronavirus: Ohio’s cases top 10,000 as 75% of inmates ill with virus at Marion state prison". Similar to the 10% of all Ohio cases being in nursing homes, they write, "With testing of every inmate near complete, an additional 692 cases among prisoners were reported at Marion Correctional Institution to raise its total to 1,057 -- about 10% of all cases reported in Ohio since March 9." They also report, "Only 381 of the prison’s 1,438 inmates do not have the virus. A total of 108 prison employees have contracted COVID-19 and one has died." [3]
I’d say they are well on their way toward achieving “herd immunity” at that prison! This virus seemingly cannot be stopped!
Are our “social distancing” actions having any affect?
News out of California seems to be making it look like our STAY-AT-HOME actions are way behind the eight ball, as we read in the Apr 17th Los Angeles Times where they describe a study in which residents were tested for antibodies. The presence of antibodies in one’s blood indicates that one has already had the virus. They report that they are getting indications that their true case numbers are likely 50 to 85 times greater than their official number of cases! [4]
For an interesting perspective on the pandemic, you may find this interview of Professor Knut Wittkowski of interest. Professor Wittkowski was the head of The Rockefeller University's Department of Biostatistics, Epidemiology, and Research Design for 20 years. He says that social distancing and lockdown is the absolutely worst way to deal with an airborne respiratory virus. You can hear his perspective here...
Here are my "back-of-the-envelope" numbers for Fri 4/17, as viewed shortly after 2:00 p.m. that day...
Friday 4/17/20 (6th doubling)
Unofficial infection estimate = 6,400,000 (100,000x2x2x2x2x2x2)
Estimated cases = 28,848 (see Note 1...357x2x2x2x2x2x2)
Official Ohio cases = 9,107
Official Ohio deaths = 418
You may recall that my “Unofficial infection estimate” was based upon Dr Acton’s remarks on March 12th, wherein she said that they estimated that 1% of Ohio’s population, or more than 100,000 people, were likely already carrying the virus around. If unabated, at the every six day doubling rate, that would mean that over half of Ohioans would be affected by now. Hmmm…considering how this virus has evaded restrictions at Ohio’s Nursing Homes and prisons, maybe half of us have already been affected, and many had absolutely no idea!
As you may recall from my 4/10 post, it was on April 4th that the modelers were forecasting that us Ohioans would be racking up to 10,000 new cases PER DAY come mid-April. Yet here we are in mid-April, and we've only reached around 10,000 total cases thus far. I'd say they (thankfully) missed the mark...by orders of magnitude.
Note 1: In my wee hour 4/12 post from this past Sunday morning I accidentally put in a 6th doubling of the Estimated cases. Thus, that 28,848 should've been 11,424 for those 4/11 numbers.
Note 2: Also note that if one is to look back and compare what I recorded for official numbers, and compare them to the "dashboard" on Ohio's coronavirus webpage, one may notice that those numbers will look different. That is obviously because the numbers they publish can change. For instance, let's say it is a Friday, and they receive info on cases that occurred earlier that week...obviously they have to adjust the numbers.
[1] https://www.dispatch.com/news/20200416/coronavirus-more-than-820-cases-and-at-least-88-deaths-in-ohio-nursing-homes
[2] https://www.dispatch.com/news/20200312/coronavirus-response-state-employees-working-from-home-visits-to-prisons-and-mental-hospitals-banned
[3] https://www.dispatch.com/news/20200418/coronavirus-ohiorsquos-cases-top-10000-as-75-of-inmates-ill-with-virus-at-marion-state-prison
[4] https://www.latimes.com/california/story/2020-04-17/coronavirus-antibodies-study-santa-clara-county