Coronavirus: Big differences in Minnesota model, national model

Minnesota leaders are confident in the model they're developing.

Differences in models that try to forecast what will happen during the novel coronavirus outbreak have left people wondering when a significant spike in cases and hospitalizations will hit Minnesota. 

None of the models are bulletproof, but Minnesota leaders remain confident that their modeling has more granular, Minnesota-specific information that has given them a level of confidence higher than the models that are the focus of the national government. 

"First and foremost, the modeling that's done is is always just that, there are going to be variables that we couldn't count on," Gov. Tim Walz said Tuesday. 

No matter where you look, Walz says model trends suggest that there will be a "slow build-up and then a rapid ramp-up," although mitigation and suppression efforts, like the stay-at-home order currently in place in Minnesota, can help delay the onset of the peak of the epidemic. 

A national model, developed by a group founded by the Bill and Melinda Gates Foundation, predicts that the peak of the outbreak will hit Minnesota around April 23, about a month before the earliest onset of the peak forecast by the models being developed by the Minnesota Department of Health and University of Minnesota. 

That model is based on morality rates, not infection rates. 

So which model should Minnesotans believe? Parts of both, probably. 

"The confidence levels, if you will, or the bands around the estimates were plus or minus a couple of weeks on either side for when that peak might be hit," said MDH Commissioner Jan Malcolm. "That range goes from the middle of May to the middle of June for what we would be expecting with peak hospitalizations. Hitting the capacity of intensive-care units was modeled out to be a couple of weeks earlier than that, so early May to early June."

A key point is that Minnesota's model has not be re-simulated since Gov. Walz introduced the stay-at-home order last week. Since then, healthcare facilities have been working to convert regular hospital beds to ICU-caliber beds, which will help save lives as more and more people fall severely ill with COVID-19. 

"We are right now just starting to be able to compare the rate of growth in our cases to what we've seen in other in other communities, so it's possible that that peak may have moved out further in time, but what we were modeling a week ago was a peak of infections being early May to early June and peak hospitalizations being middle of May to middle of June," Malcolm reiterated. 

Sign up for our BREAKING NEWS newsletters

Models have big differences

It's a morbid reality that has become all too real as the virus has killed more than 1,700 people in New York, which is far and away the hardest-hit part of America. Nationwide, White House officials estimated Tuesday that between 100,000 and 240,000 Americans could die during the pandemic. 

But the White House coronavirus task force team isn't incorporating Minnesota-specific data, and as Walz noted Tuesday, the Gates model has "holes" in it. For example, it doesn't recognize that Minnesota has closed non-essential businesses. 

Another massive difference was televised on CNN Tuesday night, as Illinois Gov. JB Pritzker said that he's estimating Illinois will need 4,000 ventilators. Pritzker said he spoke with Vice President Mike Pence and Pence said that the White House experts are forecasting a need for only 1,400 ICU beds in Illinois. 

The national model projects a need for 1,399 ICU beds in Illinois and just over 1,100 ventilators. 

"Well I hope he's right," Pritzker said, adding that his "great concern" is that "we’ll run out of ventilators, and then there’ll be people dying."

Minnesota's goal is to have 2,000 ventilators. The Gates model says the need will only be around 450. 

The Gates model is also fluid, changing totals daily. On Monday, it projected 1,800 deaths in Minnesota during the outbreak, with a peak of 70 people in one day. Those numbers, as of Wednesday morning, have dipped to a peak of 38 deaths per day and just over 1,000 deaths statewide, although it has a death range of 440 to 3,574. 

How many could die in Minnesota based on the U of M/MDH model? 

The Pioneer Press' Dave Orrick spoke with a member of the team behind the model and learned that mitigation could save approximately 20,000 lives, but the death toll could still be in excess of 50,000. 

It's a shocking number, but one that Orrick reported shouldn't be taken at full face value because the model cannot fully grasp what is happening in real life. Nonetheless, the model that Minnesota leaders are confident in is deadlier than the national model. 

"We think our model is pretty solid," said Malcolm. 

To prepare for the surge, regardless of when it happens, state leaders are identifying and preparing seven sites in greater Minnesota that will serve as alternates to hospitals, with 250 beds for non-critical patients at each facility. Another 1,000 beds will be constructed at alternate care facilities in the Twin Cities metro area. 

The Minnesota model is expected to re-simulate projections within a couple of weeks, at which point new estimates could be provided to the public. 

Next Up