Contents
Interpreting the US Graphs
Introduction
It’s probably time to do a little bit of an explanation of what is on some of the graphs I’ve created. Some should be obvious; others are a bit nuanced. I’ll use New York as an example. The headers are the names of the tabs, in order, L to R, at the time of this writing.
Early Indicators
One way to think about the first few tabs are graphs superimposed on the background of the percentages of different variants. I pull the data from CoVariants, and have used the same colors that are used there, although it looks like I may have miscopied the hex code for the color of one variant, which I will try to remember to update when I have to redo this entire file set again. The percentages for everything but the wastewater are on the left y-axis.
I placed this tab first because it has the earliest indicators of COVID cases climbing in each state. For some states, some of this data may be missing, such as how the positivity data for NY abruptly ends a few months ago. My thinking was that this is probably what most people want to know, especially if they need to travel somewhere.
I split the percentage of COVID ED visits into two separate lines. The solid red line is the real percentage. However, since that is a small proportion of all ED visits, it’s pretty dampened out and hard to really see the curve, which is much more important. I created the red dotted line by multiplying that number by ten. By doing so, one can see how both wastewater and positivity are also good indicators of COVID in the community.
Var-Sym
Same variant background, but this adds Google Search trends compiled by the Delphi Group at Carnegie Mellon. I made my own categories to make them fit in the legend better, but full details of the terms they used can be found on their site.
Var-Peds Admits
Same variant background but with pediatric COVID admissions in the foreground. The black line is the raw data, the red line is where I used a formula of reported number/percentage reporting, which is the same method for all of the following graphs that show adjustments. You can see where these two lines diverge when mandatory reporting ended in May.
Var-Adult Admits
Exactly the same as the prior, but using adult COVID admissions instead.
% ED Visits
Hopefully this is pretty self-explanatory showing the percentages of ED visits for COVID, influenza, and RSV. Note that COVID is obviously not seasonal. The colors have some transparency to show the curves instead of them being hidden behind a larger one in front.
% ED Visits (stacked)
This is the same data as the previous graph, but here each is stacked on the other to show the burden of respiratory disease in the EDs of a state.
Flu
The red line is influenza admissions and the scale is on the left y-axis. The yellowish areas are influenza ICU patients, and the blues are influenza hospitalized patients, both on the right y-axis. The darker areas of each are the numbers after doing the reporting adjustment calculation described earlier.
C19
This is the same methodology as the prior Flu tab, but with COVID instead and using different colors.
IP Bed Util
This may look complicated, but it’s very simple. The left y-axis is the number of hospital beds and inpatients. The lightest green are the available beds, the darker green are the total occupied hospital beds. Yellow and red are the the number of COVID and influenza patients.
The dotted lines are the percentages of hospitals reporting this data as given on the right y-axis. You can see the big impact when CDC stopped requiring reporting.
Adj IP Bed Util
This is the same as the previous graph, but uses the dotted lines to do the same adjustment previously described to adjust for reduced reporting. At the far right, you can see that the data doesn’t always align in each state. That has to do with exceptionally small or exceptionally large hospitals dropping out of reporting. When that happens, the adjusted numbers can get over-or under-inflated. At one time there was a way to adjust for this. CDC used to provide hospital level data for all hospitals reporting, but that’s gone now too.
ICU Bed Util
This is the same as IP Bed Util (two tabs back), but ICU beds instead.
Adj ICU Bed Util
This is the same as Adj IP Bed Util (two tabs back), but ICU beds instead.
How I Got Here
I thought some people might be curious as to how I wound up doing all of this work which until this year, had been completely on my own time, so I thought I would share that story.
This actually goes back to 2004. Besides doing hospital infection control and epidemiology, I had started speaking at conferences. Even though I’m an introvert, I found I enjoyed public speaking. This realization happened in 2001 when I took an elective course designed for dual PhD/DVM students while I was in grad school. The course was the Epidemiology of Zoonotic Disease. It was designed as a practicum, so instead of having the teachers teach it, the students had to teach it. I found I really enjoyed it and aced the course. That’s when I realized I wanted to do some public speaking. I had HATED speech in undergrad.
A professional colleague of mine asked me to speak at a regional conference on H5N1 and pandemics. One week before doing so, I was at the national conference for my professional association (APIC) and was at a breakfast. I told people at my table I needed to leave to finish preparing that presentation. A few asked if they could see my slides since H5N1 was a hot topic at the moment. That led to a number of speaking engagements around the country on pandemics, including keynote presentations, even at national conferences.
One conference was put together by the University of North Texas Health Science Center. It was primarily focused on pandemics and had some big names in the field, so I was a bit apprehensive. The university compiled all of the comments they received on each speaker and sent them to us. This is the top part of one of eight similar pages of comments I received. This is when I realized that I had developed street cred on pandemics.
Eventually I wound up leading pandemic tabletop exercises around the world. I was mostly working with either clinical or pharmaceutical leadership executives. Even before COVID hit, I wound up on a state pandemic ethics committee as well as developing the pandemic plans for a state’s primary care association.
In the fall of 2019, I had taken Al Gore’s Climate Reality Leadership Course. My intention that winter was to develop a conference presentation on the health impacts of climate change.
In February, everything changed. I was deployed as part of my federal disaster medicine role to be part of the mass quarantine efforts at Travis Air Force base in California. There are plenty of stories in the media on this. This is the link to the first one that came up. The people we were to serve were primarily State Department evacuees and their families from China.
A couple of days into this, I became symptomatic. Obviously, that made me very concerned. I was taken to a local ER and likely had one of the very first COVID tests given in the US, back when they probed to what felt like the back of the occiput. Fortunately (?), my influenza B test came back positive, so I was put in isolation and my entire team given Tamiflu. It was a scare though until that diagnosis.
I started trying to find data to get an idea of what was happening around the country. One of the first places that aggregate data was easily available in a usable form was at Johns Hopkins. I started pulling together case and death data for each state from there and plotting them together and posting them on Facebook.
After my deployment, I tried to gather as much information as possible on the disease since I had a very keen awareness of the much larger social impacts of a pandemic.
It was also about this time that I used two published sources and US Census data to determine the disease burden from COVID. A friend of mine is the CFO at a Level 1 Trauma center in my area and mentioned she wished she knew how it would affect them specifically. At the time, there was really good county level data and I told her I could probably figure out how to do that if I knew the draw area for the hospital She gave me the counties and I did just that.
From that, I realized that a number of hospitals and public health departments might benefit from a similar tool, so I built that out to where they could put in the parameter of an infection rate and the counties that they wanted the aggregated data and it would provide severity by age group for the selection. It was used widely across the US.
In March, I came across a Facebook group called Dr. Frank Models. He claimed to be a PhD chemist (true) with expertise in modeling outbreaks and pandemics (false). I realized within a week he was really mispresenting the data. When I commented on this, I got booted out of the group. I got back in, gave more evidence, and got booted again.
To counter his misinformation, I started another group called NOT Dr. Frank Models, which I knew would come up in searches at that time. I hoped to get people real information instead of his lies. He grew that group to around 50,000 members, but it eventually got banned after repeated posts of misinformation.
Now that his COVID grift is up, he moved on to election denialism. He’s Mike Lindell’s “scientist” with the “8th degree polynomial” that he claims proves election fraud. I won’t go further down that rabbit trail of his grift, but there is plenty of evidence countering his claims on both here.
I continued posting graphs on Facebook, but then realized that perhaps I should put them somewhere where someone didn’t have to be using a service. That’s what led to further development of this website, which really had just started as a blog which I used to answer questions about the Ebola case in Texas.
Since I had been able to pull together all of this county level data, I was also able to provide graphs of COVID in metropolitan areas. This came to the attention of the Chief Medical Officer of a Fortune 5 company who asked me to meet with him a couple of times a week (virtually, of course) to discuss what was going on with the pandemic.
What I never expected was the hate, vitriol, physical threats, and death threats that would come my way from trying to keep people from harm. People accused me of making this political. I believe I was the first person to identify the impacts of the politicization of the virus. If I truly were doing so, I would have just let the virus take its course without trying to intervene with Republicans, who died at a much higher rate from COVID than Democrats, other than during the original wave. That is because it hit dense population centers like NYC first.
It’s worth noting that 45 called coronavirus a big hoax at the very outset. He is a large part of the reason that it has remained politicized until this day and why Republicans are far less likely to get vaccinated and use respiratory protection. That hasn’t played out well for them.
Trump had also commented about how this was only a problem for large cities at the outset. I am now convinced that his initial plan was to let it go unchecked in large, dense population areas in order to harm more democrats for political gain. That is evil to its core.
When I started putting all of this data visualization together, I thought for sure that CDC, a university, or a foundation of some sort would start doing this type of work within a year or so. I hate the fact that a few years into this it’s just a few individuals for the most part who are trying to get good information to the public. I would much rather spend my time doing other things, but I know that there is a need and I will continue to try to fill it as long as I can.
