Washington COVID-19 model based on different assumptions
(FOX 9) - Minnesota officials have been under pressure to explain why the model the state is relying on for the spread of COVID-19 differs so dramatically from one developed by the Institute of Health Metrics and Evaluation (IHME) at the University of Washington.
FOX 9 Investigator Tom Lyden talked with one of the scientists behind that model, Professor Ali Mokdad, who has worked at the Centers for Disease Control and Prevention (CDC) and has studied the outcome of malaria programs for the Bill & Melinda Gates Foundation.
“We shouldn’t be pitting models against each other,” said Mokdad, who believes it is the broader trend lines that are important.
“At this point, all the models are telling us exactly the same story, social distancing is working and we should stay at home,” said Mokdad.
The IHME model has been showing a reduction in the number of fatalities that Minnesota could expect. As of yesterday, it showed 456 deaths in Minnesota with a peak around April 26. Just two weeks ago, the model showed more than a thousand deaths in Minnesota.
It is in stark contrast with what Governor Tim Walz revealed yesterday about a second run of the Minnesota model developed with the University of Minnesota School of Public Health and the Minnesota Department of Health (MDH).
The latest projections from that model showed a best case scenario of 6,000 deaths, and a moderate scenario of more than 20,000.
Professor Mokdad said the two models are based on different assumptions.
“Other models have been using what we call susceptible infected and recovered, and because we don’t have enough testing right now to determine who was infected, it’s very hard for us to count on these models,” said Mokdad.
Instead, the IHME model looks at the actual number of people who have died from COVID-19.
“We took the approach that we would model mortality,” said Mokdad. “And if you follow mortality and project the number of deaths you can do the backwards calculation.”
He said the model is following the trends perfectly and the confidence level is improving as it is updated daily with more data.
The IHME model does not contain data specific to Minnesota, and runs only through August. The University of Minnesota model is loaded with specific Minnesota demographic and health data and goes out 12 months.