Tomorrow it appears that the Trump administration will discuss the models they've been using to predict the course of the COVID-19 pandemic in the United States.
But a Washington Post story about the coronavirus modeling says that a University of Washington researcher, Chris Murray, independently has been preparing forecasts for both the United States as a whole and each state that came up with similar results as the Trump administration model.
Debbie Birx, the coordinator of the White House coronavirus task force, said that federal scientists reviewed a dozen different models for how this pandemic might play out before constructing their own. Using techniques learned from tracking HIV, tuberculosis and malaria, she explained, “we went back to the drawing board over the last week or two and worked from the ground up.”
Simultaneously, the Institute for Health Metrics and Evaluation at the University of Washington, led by professor Chris Murray, was doing something similar. “When we finished, the other group that was working in parallel, which we didn't know about, IHME and Chris Murray ended up at the same numbers,” Birx said in the Rose Garden on Sunday evening.
...The government is keeping its own projections private, but Birx twice said that people can study Murray’s work to better understand White House considerations. “If you go on his website, you can see the concern that we had with the growing number of potential fatalities,” she told reporters.
Naturally I was eager to check out the website maintained by Murray and his team at the Institute for Health Metrics and Evaluation (IHME). While COVID-19 is ugly, the website is a thing of beauty. Clean, clear, easy to understand. (With one exception that I'll talk about below.) And it's going to be updated daily.
Here's the forecast for Oregon deaths. The solid line is actual deaths; the dotted line is the projection. You can see that the death projection becomes almost flat by mid-June, which is good news.
However, the forecast has a large 95% UI (Uncertainty Interval, which is the same as a Confidence Interval). That's the light brown shading on either side of the dotted line. This is what I didn't understand at first, until I clicked on some of the dates and saw that, for June 15 as an example, the projection of 466 deaths by then had a range of 112 to 868.
I'm assuming this range reflects how strongly Oregon engages in efforts to mitigate the spread of the coronavirus. A web page describing the methodology seems to indicate this, though it isn't crystal clear from my admittedly brief reading of the IHME site.
The estimated excess demand on hospital systems is predicated on the enactment of social distancing measures in all states that have not done so already within the next week and maintenance of these measures throughout the epidemic, emphasizing the importance of implementing, enforcing, and maintaining these measures to mitigate hospital system overload and prevent deaths.
Here's more good news for Oregon from the IHME model.
The forecast shows that our state will have enough ICU beds to meet the COVID-19 demand on the peak resource use of May 3, along with enough hospital beds under even the worst projection. But the caveat is, obviously, that ICU and other hospital beds also are needed for patients with different problems. So it's good that our hospitals are planning ways to handle non-COVID 19 patients in various ways.
By contrast, here's the scary forecast for the state of New York. Yikes! Peak hospital resource use is soon, April 9, when 11,070 ICU beds would be needed, with just 718 beds available, producing a shortage of 10,352 beds. I don't know if this takes into account the hospital ship that has been sent to New York, or the hospital bed capacity being constructed as rapidly as possible.
Lastly, here's the total deaths forecast for the nation as a whole. The peak number of deaths is on April 15, after which the curve begins to flatten out. The 95% UI for August 1 is between 39,174 and 141,995 deaths, with 82,141 deaths being the projected number. All of those numbers are super-scary.
That said, I was strangely reassured by poking around the COVID-19 projections on the IHME web site. There's something about solid scientific analysis that is grounding in a time of dangerous uncertainty. The human misery is absolutely real and can't be ignored.
However, I felt better knowing that researchers like Chris Murray are working hard to provide policy makers at the local, state, and federal levels with the best possible information needed to get us through the coronavirus crisis.
I just wish there was a toilet paper availability forecast so I'd know whether I could modify my "two squares per poop" rationing sometime soon.