Expert explains why models of Virginia’s coronavirus peak keeps changing

Posted at 7:33 PM, Apr 07, 2020
and last updated 2020-04-07 20:25:25-04

RICHMOND, Va. -- A widely followed model predicted the peak of hospitalizations in Virginia related to COVID-19 earlier than previously reported.

Virginia's COVID-19 peak will hit in late April, according to the latest model from The University of Washington’s Institute for Health Metrics and Evaluation.

Just last week, the model predicted that Virginia's peak wouldn't be until late May into early June.

During his Monday press conference, Governor Ralph Northam cautioned that various models “change literally everyday depending on the data that’s put into them.”

"We will certainly continue to follow the trends. And you know, if and when we need to make adjustments and in our guidelines, we will certainly do that," Northam said.

The IHME model has predicted that April 22 will be Virginia’s deadliest day for COVID-19 related deaths.

Dr. Danny Avula, the Health Director for Richmond and Henrico, said there are numerous models being created by states and individual health systems.

For instance, Avula said Bon Secours Health System collects data and analyzes trends and cases throughout their footprint in four states, including Virginia.

“I’m cautious about the significant shift that we saw in the last 24 hours in some of those models. I think they were really a result of a dip in cases and death reporting in New York" Avula said.

In a press release, IHME researchers contributed new data regarding social distancing and its impacts on European countries for the projection changes.

“The time from implementation of social distancing to the peak of the epidemic in the Italy and Spain location is shorter than what was observed in Wuhan. As a result, in several states in the US, today’s updates show an earlier predicted date of peak daily deaths, even though at the national level the change is not very pronounced,” according to the release.

Peter Aiken, a professor of information systems at Virginia Commonwealth University, quoted British statistician George Box when asked about the discrepancies in the models.

“All models are wrong, but some are useful,” Aiken explained.

“We still don’t know whats going to happen in the future, but we can hope the models will tell us some useful information. We will take that useful information and put it to good use for everybody,” he stated.

Models serve as a forecast or best guess to help predict what will happen in the future.

“We are looking at about what’s happening about two weeks ago,” Aiken explained. “So, it takes that long to process the information and to understand it.”

State health officials said researchers with the University of Virginia will soon release their own model specific to the Commonwealth. The model is scheduled to be made public by the end of the week.

Philip Bourne, a professor and dean of UVA’s School of Data Science, said the school is also working on other models related to the pandemic.

Researchers are identifying medications that may be repurposed to combat COVID-19 and additional pandemic modeling to help flatten the curve.

“The sciences are coming together in an unprecedented way,” Bourne explained. “It’s not the usual way to do science, which is competitive. It’s way more cooperative in the more than 40 years of my scientific career than I’ve ever seen.”

Despite the models, Northam urged Virginians to continue social distancing, hand washing, and staying home.

“That’s what science tell us works,” he said.