New model predicts Minnesota's infection peak will arrive earlier, be less severe

A new model of the coronavirus spread in the U.S. predicts the peak in Minnesota won’t be as severe as officials feared, but it will hit earlier than they had hoped.

The model was developed by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington, a center founded by the Bill and Melinda Gates Foundation.

“The way we did our model was different than anyone else,” said Professor Ali Mokdad, one of several experts involved in the mathematic model. “We modeled mortality.”

Other models have based their analysis on infection rates, which are difficult to predict because of inadequate testing. The IHME model looked at mortality rates, per capita, over a period of time.

“We don’t know how many are positive. So, let’s go with mortality and we’ll reverse engineer infections to decide how many ventilators and beds we need,” said Prof. Mokdad in a Skype interview.

That model predicts April will be the cruelest month of all. The slope of the curve skyrockets upward near the beginning of the month, with the peak coming around April 19, with an estimated 49 deaths in a single day, and a total of 1,039 deaths. 

In the IHME model, hospital admissions closely follow that curve, hitting peak capacity the day before with a projected demand for 4,557 hospital beds and 712 ICU (intensive care unit) beds.  

As of Sunday, Minnesota currently has 2,621 available hospital beds and 255 ICU beds. Health officials say that number can be augmented with 1,400 available surgical beds. The state is also setting up secondary sites across the state to care for the less critically ill, with 2,700 beds. 

The IHME model shows the peak of the pandemic coming two months earlier than a different mathematical model developed by the Minnesota Department of Health (MDH) with the University of Minnesota School of Public Health, which Governor Tim Walz and state officials have said is guiding their decision making. 

That model included two scenarios. The first showed what would happen without any intervention, and the results were dire. That worst case scenario showed the peak of the pandemic coming in late May, with more than two million infected, and a potential death toll of 74,000.

But a second scenario, with interventions like the two-week stay at home order and continued social distancing, show Minnesota would hit its ICU capacity around June 7, and the peak of the pandemic would occur later, June 28 -- three dramatically different scenarios in three different months.

Governor Walz urged people not to focus on specific dates or numbers but, instead, the general trend lines.

“I’m certainly not dodging the specifics, but it sets up false expectations for accuracy,” said Gov. Walz Tuesday in a conference call with reporters. 

Walz admitted the peak of the pandemic may have moved by two weeks or more, but said, “If there is a path to err on, it’s the side of worst case. We need to be ready in a few weeks.”

Minnesota Health Commissioner Jan Malcolm said she was aware of the IHME analysis, but said she preferred the model developed by MDH and the University of Minnesota because it has specific state demographic data.  

Malcolm said there are plans to re-run the mathematical model with new data.

“When you look at the health profile of Minnesota, you do better than other states,” admitted Prof. Mokdad.  

Other models have used a case fatality rate (CFR), that calculates deaths as a proportion of the number of infections. That is similar, but different than a mortality rate that does not rely on a prediction of infections.  

Prof. Mokdad said the IHME model does take into account mitigation efforts, like social distancing and stay at home orders, which he said are extremely effective. He said it is unclear if the virus will come back later or stay during warm weather.

“We are on a trajectory, but we can do something about it,” said Prof. Mokdad. 

“The message is for all of us to stay home and not lose the ones we love.”