The case against “cases”

Years ago in the news business, a measure of success in the digital world was page views. We all wanted them and measured success based on how many we got.

The page view obsession eventually wore out as we realized this: It was kind of a phony measurement, mainly because we could drive that number with sensational or even salacious content that would appeal to non-local audiences. That in turn, brought traffic to advertisers that didn’t really do the businesses any good.

For example, if in Tallahassee we ran across the story of a three-headed alligator that was eating family pet kittens, that would draw readers from across the globe. But if I’m a local bank or furniture store buying ads on the web site, what good do those eyeballs do for my business?

So we moved on to other more relevant metrics, which wasn’t to say we ignored page views. Local page views are great! But we also wanted to see how long readers stayed on site and how often they returned and stuff like that. That is, we wanted the most relevant info we could find.

I’m suggesting today that COVID “cases” are the page views of the pandemic.

Again, that’s not to say they are irrelevant, but they are not the most useful metric to measure the impact of this deadly, dreaded virus.

This is an important issue to consider as schools start back up and with the backdrop of our local university settings.

Here are the reasons I don’t think cases are the best measure of the impact of the virus:

  1. Not all cases are created equal
  2. Contracting the virus isn’t the worst case — being hospitalized or dying from it is
  3. Cases are tied to testing which we know is imperfect, inconsistently administered and evolving.

Let’s start with “not all cases are created equal.”

According to the Florida Dept. of Health, the state has had just under 50,000 children (ages 17 and under) diagnosed with COVID-19, through Aug. 26. Florida, also as of the 26th, had about 659,000 total cases. That’s 7.6 percent of cases. But under 18-year-olds make up almost 20 percent of the population, according to the U.S. census, so clearly COVID is weighted away from young people in terms of number of cases (at least the ones that have been tested).

But what about severity?

Let’s compare young people to the oldest segments of our population.

And let’s go national so we don’t get any demographic wonkiness Florida might add to the mix.

In the U.S., those 17 and under have made up 8.1 percent of the cases, while those 75 and over have made up almost the exact same amount — 7.9 percent. That’s according to the CDC.

But when it comes to the share of fatalities, a total of 85 Americans in the under 17 age bracket have died — out of the 150,000-plus total — that’s six hundreths of one percent of the total fatalities. For those over 75, almost 79,000 have died, which is 58 percent of the fatalities.

(And let’s just stop right there and say that some of the “they were going to die soon anyway” rationalizations for all those elderly deaths are disgusting. Even if an elderly life is cut short by three months, it’s a tragedy. I’m not diminishing elderly deaths AT ALL. Rather, I am showing how the threat to the elderly compares and how measuring cases can distort the view of how the pandemic is affecting any given community).

Let’s put those numbers another way.

If you divide the number of fatalities into the number of cases for those 75 and older, you get a fatality every 4.3 cases, or a 23 percent fatality rate for cases. For those under 17, it’s a fatality every 4,100 cases or a .02 percent fatality rate.

Here’s the number I’ve been trying to get to. By dividing 4.3 into 4,1000, I get it:

It takes 953 cases in someone under 17 to equal just ONE case in someone 75 or older, when it comes to the risk of fatality.

So, when you see that Florida added X thousand cases in a day — what was the median age of those cases?

One reason Leon County’s fatality rate has been so low is that the median age of a case is 29. Compare that to Charlotte County (lots of retirees) where the median age of a case is 54. Charlotte County has had 110 deaths out of about 2,600 cases. Leon: 33 deaths out of about 6,200 cases.

Here’s the other thing I don’t think we fully recognize and grasp as a society.

Lots and lots of people die every year and more of them are young than we would ever want to imagine.

Here is a chart of U.S. annual deaths for 2020 (actually, starts Feb. 1, which is I guess when COVID data starting flowing, but you get the idea) from COVID and all other causes from the CDC.

I guess I’ve never really thought about the fact that millions of people die every year in America.

What’s really surprising is the 15,000 or so people ages 14 and under who have died this year of all causes. As you can see, COVID victims make up just .3 percent of all deaths in that age range. Even in those over 85, COVID has made up less than 10 percent of all fatalities. (Yes, I know there are disputes about the data and how things are coded — I hear from both sides that the numbers are fudged higher or lower to help one side of the other. Frankly, I don’t think either side is smart enough to pull off that kind of a conspiracy).

So, what’s my point?

First of all, looking at cases is one of the worst metrics to use when analyzing the impact of the virus.

Hospitalizations are a clear signal of a severe case and, of course, death is the ultimate price to pay for contracting COVID.

One telling fact here is from the University of Alabama, which has garnered national headlines recently for COVID outbreaks. More than 1,000 University of Alabama students have COVID-19.

Of those, do you know how many have been hospitalized? Zero. Now, given the delay from the onset of symptoms to hospitalization, that number may or may not start moving.

As an aside, here is a question: Why in the world would you want to send infected college students home, where they interact with all sorts of more vulnerable people while traveling and then while back in their home communities? In a dorm, they are just passing around COVID to a group of people with just a tiny risk of having a severe impact from the disease (I had somebody in health care tell me recently that FSU’s campus will be the safest place in the world by November because it will have achieved its own little “herd immunity”).

This is also why five Notre Dame professors penned a public letter urging the administration to keep on-campus learning, which the university ultimately did.

Second, when looking at the big picture, the risk for young people when it comes to COVID remains very, very small.

Here are the 10 leading causes of death for young people, from a New England Journal of Medicine study in 2018.

Even if COVID death numbers for this age group (19 and under) more than doubles the rest of this year to almost 200, COVID-19 wouldn’t even make the top 10 causes of death for young people (chronic lower respiratory disease took 274 lives in 2018).

On the other hand (a clause you will read a lot on this site), there is still much we don’t know about how the virus affects people, including the possibility of myocarditis (weakness of heart muscle) and possible neurological impacts.

My third point — if you think this is an attempt to oversimplify the issue of attending school or the like, I refer you back to earlier blog postings.

This is still incredibly complicated.

There is still the issue of young people infecting older more vulnerable people. There are still tough questions about how students interact with teachers, school staff and others on campus. And even within the young people cohort are individuals who are at higher risk than their peers, because of underlying health conditions or other factors.

No, this isn’t simple. But that cuts both ways. Merely pointing at case numbers and declaring that being in school is unsafe is also oversimplifying the situation.

And on the other end of the spectrum — what are we doing to protect those elderly Americans who are at such a scarily vulnerable place with COVID?

Finally, you’ll notice there is no politics in here — there are plenty of places you can go to have that conversation. I’m just interested in data and what conclusions and inferences can be drawn from that data. And by the way, lest you think I’m a complete idiot, I’ve had this piece and those like it reviewed by a journalist, two health care professionals as well as the toughest judges of all — family members — pre-posting.

Feel free to chime in with your views, of course. But if you are going to take this down a political path, you are unlikely to get engagement from me on that — I’ll probably only give you a page view.

On the other hand ….

A dive into COVID data.

One thing very frustrating to me about the current political climate is that those who have sold out to politics view all data through that lens. When it comes to COVID-19, that means “sold out” Republicans see only that data that minimizes the impact of the virus. And “sold out” Democrats only see or retain data that affirms their view that COVID is cataclysmic and is the fault of Republican leadership. Guess what? There is a ton of data out there which is contrary, textured, conflicting and nuanced. And since I’m neither an R or a D, the scales aren’t covering up my eyes when it comes to this rich, fascinating (and obviously sad and sometimes scary) data.

The point of this piece will to find examples of data that isn’t fitting one narrative, but that has caught my eye. Some of it (like “Manic Tuesdays”) will not really really be supportive of either side (doesn’t sound ridiculous when talking about a pandemic?) but will simply be informative.

I will do my best to include source material on everything I post. If you see a source that is questionable or an error I have made, PLEASE let me know so I can correct it. Hopefully, if something “on one hand” rubs you wrong, what’s “on the other hand” will be comforting.

Just another Manic Tuesday ….

One key thing to remember is that statistics are reported unevenly through the week. Nate Silver of has done a good job pointing this out. Day-to-day comparisons (“deaths were up Tuesday over Monday) are useless because weekend reporting is not as complete. You can see from this chart below, taken from that there are 5 days of higher reports, then two days of lower. In fact, Tuesdays tend to be the highest reporting days because of weekend “catchup.” CONCLUSION: Don’t get overly excited when weekend/Monday number or low, or overly concerned if Tuesday numbers are high. The best way to compare is week over week (this Monday over last Monday).

Florida cases vs. hospitalizations

The increase in Florida cases has clearly been marked and alarming. I understand that there should be some delay between a soaring case count and fatalities. What is less clear to me is why hospitalizations have not ramped up — at least on a statewide basis — since the increase in cases.

One site I’ve been checking is this AHCA spreadsheet showing ICU bed vacancy. When I first started it checking it, a week or two ago, the vacancy rate was 19 percent.

This is a screengrab from the morning of July 6 (when I wrote this section).

You can see that the ICU bed availability is 21.55% — in other words, it hasn’t gone up since the surge in cases. Over the past few days that number has fluctuated from between 19 and 22 percent.

For context, when I asked TMH CEO Mark O’Bryant what the typical ICU vacancy was in NON-pandemic times, he said 20-30 percent. So, right now, we are within that range (barely). Here is the link to that conversation, which is worth your time. The discussion of typical ICU availability starts at the 17-minute mark.

Going to school on schools

The announcement by Education Commissioner Richard Corcoran that all Florida schools were to reopen in August was quite a shock, to me at least.

But the issue of what to do about opening schools has been hotly debated among epidemiologists, pediatricians, educators, politicians and more, for many weeks.

Last week, to surprisingly little fanfare, the American Academy of Pediatrics issued a position calling for all American schools to reopen.

From the statement: “The AAP strongly advocates that all policy considerations for the coming school year should start with a goal of having students physically present in school,” reads the guidance. “The importance of in-person learning is well-documented, and there is already evidence of the negative impacts on children because of school closures in the spring of 2020.”

Other European countries have already opened schools — Denmark, Austria and Germany to name a few.

There have been a number of studies finding children are not affected by COVID at near the level of adults and especially the elderly population.

Here are a few:

Study: Children half as likely to be affected (Washington Post)

French study: School children don’t spread COVID (Bloomberg)

Netherlands study: Children don’t spread COVID (Netherlands National Institute for Public Health)

(Interesting quote from the Netherlands study: “Based on source and contact tracing from the beginning of the epidemic, we see the following: looking at 10 COVID-19 patients who were <18 years old, they had 43 close contacts, and none of them became ill, whereas 8.3% (55/566) of the close contacts of the 221 patients who were ≥18 years old became ill. Now that widespread source and contact tracing is ramping up again, we will be able to update this information with recent data in summer”)

One thing to keep in mind is that, statistically speaking, children are simply not dying of COVID.

Here is a chart as of a couple weeks ago, for the entire U.S., from the CDC:

To put those 26 deaths in perspective, If you look at “all cause mortality,” about 9,300 people ages 14 and under die in a year in the U.S. (most recent data I found on this was 2017, below). Causes like accidental drowning, car crashes and the like far dwarf the threat of COVID to children.

Now, that’s not to say there is no threat (aka “on the other hand”). You may have heard of the mysterious illness in New York and other places that was thought to be COVID related that targeted children. To be honest, there hasn’t been much recent reporting on this — I’m not sure why.

Here is a story from five days ago on some new cases emerging, taking the total number to 300. (WBRC)

Here is a story from a few weeks ago on a Boston Children’s Hospital study on the disease (WBUR)

A big issue with regards to going back to school is the threat to teachers.

One issue is that teachers are often in vulnerable age groups (U.S. News and World Report)

I have not been able to find any studies that show what risk teachers face from school reopening — if you can find one, please share.

The clearest danger

Let’s make no mistake, the most vulnerable populations are the elderly and those with medical conditions (particularly lung issues) that COVID attacks.

Let’s go back to the chart I referenced above:

This image has an empty alt attribute; its file name is image-2.png

A staggering 80 percent of COVID deaths come from people over the age of 65.

I’m not that smart, but it seems to me that our society should be laser focused on protecting folks in that age group.

OK, I’m going to stop here. I will probably do a part 2 to this, but want to publish this before what I’ve written is outdated. Hope you find this useful. Please leave comments with questions. And don’t forget to subscribe!