Coronavirus Episode: How Bad Are Things in Your State? How Can You Plan?
I was on Telemedicine visits with patients the better part of yesterday and I was using the IHME model to try to predict when was the best time to see them for their procedure. The patients loved it, so I thought it would make a great blog. So how bad is your state? When will it open back up? Does it have enough hospital beds? How can you use all of this data to plan your life? Let’s dig in on how you can easily figure all of that out.
How Bad Is Your State (or Country)?
It’s always frustrating as a physician-scientist to see the media throw up an absolute number of coronavirus cases in a state as that number can be relatively meaningless in figuring out how bad a state is or isn’t. Why? Because our states have vastly different populations. In addition, they all have vastly different testing rates. Hence the numbers we should all actually be looking at is the number of cases per million and the number of tests per million.
First, for the world, I use this page and for the United States, I use this page. Let’s dive into the country data first by comparing the U.S. Italy, and Germany. For example, the US this morning has 489K cases compared to 144K for Italy and 118K for Germany. So the US is the worst off, right? Not necessarily, because the coronavirus cases per million population numbers tell a different story. For example, the US has 1,417 cases per million, Italy has 2,376, and Germany 1,411. In addition, if you look at tests per million it’s 7K for the US, double that number in Italy at 14K, and 16K for Germany. Meaning that, if the US had been testing at the same rate as Italy, it may have about the same number of cases per million. Since Germany has tested more of its citizens per million and is still clocking in at only about 1K cases per million, it’s the best off of the three countries.
This is also easy to do the same with US states:
Here, New York is the most infected per million, but it’s also done around four times as many tests as everyone else. You can see that Michigan is about twice as infected as Colorado and given that it’s done about a quarter of the number of tests as New York, for all we know, they could be equal in actual cases per million. Texas and California have the fewest cases per million, but both are not testing “superstars”.
Hospital Beds and Peak Utilization
There are a number of models being used out there to predict the peak utilization of healthcare resources. Why should you care? Because these models can help you plan your life and business!
The one that I find the most accurate is the IHME COVID-19 model from the University of Washington. Why should you care about these models for your state? It’s likely the most accurate way to predict when your state stay at home order will be lifted and when elective medical procedures can be performed again. Meaning if you own a business, you should be using this information to plan.
When using the IHME model, first pick a US state up top on the page linked above. Here I’ve done that for Colorado:
From the pop-up box seen above, Colorado has more than enough hospital and ICU beds and if I move my cursor towards the top of that hill, I can see that our peak utilization of ICU beds is happening now (April 8th-10th). While physicians can perform more urgent procedures, our restriction on elective medical procedures is set to expire on April 26th. Is that reasonable? Yes, if I move the cursor over that far, our ICUs will have about 15% COVID utilization by then. Hence, the focus of the healthcare system can move away from COVID-19.
However, if I look at Iowa, the peaks move over to the right and by the time Colorado will be allowing elective medical procedures, Iowa ICUs will be at max utilization. Hence, it’s unlikely that they’ll be performing elective medical procedures on that date, as their hospital systems will be otherwise engaged with COVID-19 cases.
Now let’s look at New York, where many of these numbers flip in a bad way:
New York is going through peak ICU bed utilization now, but they need many more ICU beds than they have. They won’t get to the point where Colorado will be on April 26th with 15% ICU beds filled with COVID-19 cases until about two weeks later.
Looking at when these places will reopen and get past their shutdowns, they will get there at DIFFERENT times. Colorado will get there sooner and Iowa and New York later. I suspect that local public health officials will be using these same models exactly the same way we’re doing this today. These models are also changing day by day, so check back often to see how things are actually playing out.
The upshot? You can use the same data that public health officials are using to make the decisions that impact your life and business. The difference is that the public health gurus in your state are less likely to tell you what’s coming up and when, as they don’t want to get this wrong. Hopefully looking at all of this data will allow many of us to feel a little bit more in control, and allow us to more accurately plan our immediate futures.