To improve predictions, Minnesota wants to learn from sketchy COVID-19 models


Minnesota is using a $17 million federal grant to learn from COVID-19 forecasting losses over the past few years and improve its predictions for the next outbreak.

Dr. R., associate professor of health informatics at the University of Minnesota. Adams Dudley said better estimates of infectious disease cases and spread could improve responses and target them to high-risk areas or populations rather than the entire state. Leading the effort.

Better forecasting could also prevent losses during the Covid pandemic, when inaccurate predictions undermined public confidence in government quarantine orders and restrictions.

“Think about the cost of getting the public trust wrong,” Dudley said.

A consortium of Minnesota agencies announced last week that it is one of 13 research groups across the United States that will receive federal funding to better prepare the country for the next public health emergency. The group hopes to make important discoveries based on Minnesota’s unique experience with COVID and a new shared medical record-keeping system that does not exist elsewhere.

Although COVID remains a concern — a revised state dashboard this month identified low but rising viral levels in wastewater — the Mayo Clinic and other major health care organizations stopped predicting its spread months ago Was. The federally supported model, called ENSEMBLE, still predicts trends in COVID hospitalizations, but not infections or deaths.

So the time is ideal to learn from the pandemic and use that experience to create more accurate tools, said Eva Enns, a U of A public health researcher who helped build Minnesota’s early COVID models in 2020.

Due to environmental conditions and global dynamics, he said, it probably won’t be another century before the next public health emergency arises.

“This is probably not going to be another COVID. This is going to be something new,” he said. “But we need to explain what happened while these details are fresh.”

Case numbers are getting better

The first step will be to survey Minnesotans to get a better understanding of their daily activities and face-to-face interactions — a key data element in calculating how a disease spreads. The mobility estimates used in the Covid forecasts were crude, and did not take into account differences based on weather or urbanization.

Once the survey is validated, it can be used monthly to detect changes in public attitudes, which may affect how quickly the pathogen spreads. Everything from political leanings to bad memories about the COVID public health response could influence whether people wear masks or stay home depending on future events.

“Will there be more resistance because, you know, [COVID] It lasted longer than people were told?” Enns said.

The research will also rely on the new Minnesota EHR Consortium, which links medical records from clinics and hospitals to identify trends. The consortium’s first project earlier this year was a near-real-time dashboard of drug-related ER visits in Hennepin County.

Minnesota was one of the first states to publicize COVID modeling projections in the spring of 2020. Governor Tim Walz issued the statewide stay-at-home order based on a prediction that 74,000 Minnesotans could die if nothing was done. Risks of death were overestimated due to unclear information about the new coronavirus strain; More than three years later, the state is now approaching 15,000 deaths.

Dudley said the modeling on the rate of deaths per COVID cases was inaccurate because the denominator of total COVID cases was too low and was based mostly on hospitalizations. Tests to identify mild infections were not initially widely available, and field reports to public health agencies were inconclusive.

“The numbers were wrong because we didn’t really have a system to know how many cases there were,” he said.

Dudley said an analysis of Minnesotans’ completely updated clinical records would solve that problem by identifying broad groups of patients who visited their doctors with the same suspicious symptoms.

The consortium can also serve as an early warning signal in case of understanding common disease patterns and alerting health officials to an increase in unusual symptoms – such as loss of smell and brain fog symptoms that may be associated with COVID. There were early symptoms of.

That solution also has limitations. COVID caused asymptomatic infections that did not show up in clinical records and further obscured the true mortality rate of the disease.

predictive tools

Consortium leaders said their solutions must be flexible. The next outbreak may involve another respiratory infection or have little in common with COVID, perhaps involving digestive problems or emerging from insects.

Despite numerical missteps, early Minnesota COVID models succeeded in their primary purpose: to give state leaders an estimate of how policies like mask mandates or nursing home closures would affect viral transmission.

Kristen Sweet, an infectious disease manager for the Minnesota Department of Health, said new prediction tools need to be developed so they can be easily understood and actually answer the questions facing decision makers during the crisis.

“What is useful to people and how do we communicate both the strengths and limitations of different analytical tools?” said Sweet, who is also a co-leader of the grant-funded partnership.

The new forecasting tools will also be tested by running simulations to see whether the tools would have identified COVID patterns sooner.

“We’re going to go back and pretend we don’t know anything more than March 2020 and see if it would have worked and how quickly it would have worked for COVID,” Dudley said.

Leave a Comment

“The Untold Story: Yung Miami’s Response to Jimmy Butler’s Advances During an NBA Playoff Game” “Unveiling the Secrets: 15 Astonishing Facts About the PGA Championship”