COVID 19 - The Coronavirus: Playing with Numbers

So I'm a math person. I like to look at numbers and look for patterns. I've been watching the numbers being reported on COVID.

I look at numbers from multiple sources, because (surprise, surprise) the numbers vary based on where you find them.

The official CDC page seems to update a little slower. I took all of these screenshots on April 23rd at 9 AM, but the data here says "As of April 21."

Fox News was more recent. It had last been updated the night before (April 22 at 7 PM).

This screen shot from CNN said it was updated as I was checking (4/23 at 9 AM).

Anecdotally, I've found CNN normally updates sooner and more frequently than Fox while the CDC tends to lag behind. I will check numbers over multiple days sometimes as the figures can change (more on that later).

Both CNN and Fox listed similar sources (like the CDC in my first screenshot). John Hopkins had worldwide tracking and they mentioned a pair of other sites like the COVID Tracking Project and World Meters.

I used a link from Fox to go to the COVID Tracking project.

Their numbers were updated as of 8:13 AM and a little different from the other sites. The deaths were lower, because they broke confirmed and pending into different categories. CNN and FOX both reported 46,785 deaths. COVID Tracking stated 4,249 pending + 42,217 deaths which equaled 46,466. 

It wasn't a large difference, but it was a difference and I thought worth mentioning.

Unlike the news sites, I found it interesting that the COVID Tracking project broke down their data on a state level, included links, and tried to grade the data (Which is subjective, but important - There have been discussions in the news lately about how 4 states started to compile test results of different types of tests into their data, but I'm not going into that here).

I typically look at Worldmeters for my daily check. I like the format and it allows me to see graphs of past numbers easily (Notice - it has a higher death count. The last updated says "April 23 12:58 GMT" which translates to 8:58 PM Eastern time)

I like how they have a timeline that includes news that may affect the quality of the data such as definitions changing over time and data being updated.

One of the reasons I like to look at the numbers myself is to help me spot bias in news articles. The numbers shouldn't change, but graphs and data can be presented in ways to lead people to different conclusions.

For example, Georgia has been mentioned on CNN's homepage multiple times as their governor wants to open back up sooner and are less supportive of the stay at home orders. This screen shot was taken May 1st according to my naming convention.

The data blurb was not a lie from what I saw (as I listened to a Dice Tower video - Screen shot also May 1st).

Mathematically though, that wasn't the fastest rate of growth on the chart. Maryland had less total cases and more growth than Georgia, but they weren't front page news. More important to me, Virginia had more new cases, but was growing quicker proportionally! Georgia's population was listed as 10.62 million in 2019 when I Googled it. Virginia's was 8.536 million (also 2019 estimate).

Yes, Georgia topping 1,000 new cases was a big jump in one day's worth of cases compared to the day before, but Virginia's rate of increase was consistently faster (and I didn't circle it, but look at Maryland.

I'm not an expert. I do want to hear from experts. I'm interested in if people see things I don't in the numbers, and then I look to see how they came to their conclusions. I don't want to be led astray and chase knee-jerk reactions even if they are true in the moment. I want to know the moment and have an idea of how it fits in the larger context. 

The world is too much for me to track as an armchair researcher. It's good to see bits of good news here and there, though!

Yeah, I went there. It's good to keep a healthy sense of humor.

I recently found out Virginia has it's own website with official numbers. I looked up my county right away.

For perspective:

I looked up Charlottesville, too, because the city and county are closely connected in many ways.

Now, I've shown you where I'm getting the numbers I look at (and maybe some insight into why), but I have to mention that there are still questions about how accurate the numbers are. There have been news articles about a large number of false positives as tests have been rushed into use (which would support someone saying real numbers should be lower). There have been news articles about large numbers of asymptomatic carriers, morgues without access to tests to determine causation, and many people being turned away from being tested early on due to not traveling (which would support someone saying real numbers should be higher).

Unfortunately, if someone were convinced beforehand of something (like COVID is a hoax or it's the plague) they can find articles to skew things towards their view. Confirmation bias is something inherent to human nature (I'm no different), so it's important to be aware of it.

I wanted to see how deaths listed as from the Coronavirus compared with other sources of death in the United States. It's been frequently compared to the flu.

I went to the CDC website and they listed the causes of death for 2017.

If it's hard to read, it says that there were 55,672 deaths in the US due to "Influenza and penumonia" in 2017. The stats list it as 2% of US deaths.

The leading cause of US deaths was "Diseases of heart" with 647,457 deaths for 23% of the total. 

I divided both numbers by 365.25 (days in a year including leap years) and rounded to tenths.

Diseases of the Heart average - 1,772,6 deaths per day

Influenza and pneumonia average - 152.4 deaths per day

I made a spreadsheet and started typing in daily amounts of deaths according to Worldmeters. I recorded from March 27 to May 15 (51 days). I picked March 27, because I felt the number of deaths started to quickly accelerate.

Minimum daily deaths - 354 on March 27
Maximum daily deaths - 2,683 on April 21
Average of time period - 1,713.8 per day
Days out of 51 that beat Diseases of the Heart Average - 24
Ending daily deaths for recording period (May 15) - 1,753

There's always a chance I mistyped a number into a calculator here or there, so feel free to double-check me. It will either prove I was right (yay!) or it will fix an error (also, yay!...just not as enthusiastic of one if I'm honest).

This is the data according to Worldmeters at the time I'm finally sitting down to write this.

Compared to the 2017 data, the Coronavirus is in 7th place as a cause of death. It ranks more deadly than diabetes to Americans, but not as deadly as Alzheimer's.

I'm sure there will be many research papers written about this in the future. There are lots of questions now and I hope many more will be answered over time.

Something I have seen mention in the news and compared to what's going on now is the Spanish flu.

I did a little bit of side research on it out of curiousity.

I'm no epidemiologist and I recommend you look into some of this on your own!

I've also written quite a bit some I'm going to wrap up with this question: Will there be a second wave of Coronavirus like with the Spanish Flu?

It's a topic that's been in the news and is a concern with easing restrictions too soon. The counterargument is if restrictions are eased to late there may be worse economic damage that leads to loss of life in others ways.

What's the answer?

I'm not an expert. All I have are my observations and opinions.

I thought that the initial reaction in Virginia was too knee-jerk and reactionary, but I supported staying at home to reduce the spread and save lives. I was sad that I wouldn't get to finish more of the school year and that I had to give up traveling to see family and friends.

When hospitals were asking to reopen for elective surgeries I supported that idea, as well. Yes, they could be a source of where the virus was, but if the hospitals needed to open up options to stay healthy themselves so that they could treat people? That was May 15th.

There has not been a massive outbreak in the millions and millions and I would like to think that's because most people have chosen to stay at home. Unless there's a parallel world I don't know if that's ever going to be proven conclusively. As I write this, restrictions are easing up. Will there be a flare in cases in three weeks? Or will things be leveled off and we won't see a second wave?

Time will tell.

Sources:

My spreadsheet I was playing on

CDC website - Fast stats on deaths  (Was updated for 2017 when I used it)

CDC website - Heart disease

World meters - Where I got daily numbers from COVID from

Worldwide numbers from John Hopkins

Map of Virginia with state government data