Countless statistical models have been used to forecast how many people might be infected by the novel coronavirus, how many might die from it, and how many hospital beds and ventilators those patients might require.
One prediction tool from researchers at Washington State University — developed in collaboration with MultiCare and the Spokane Regional Health District — has been used by health care experts locally and abroad to prepare for surges of COVID-19 cases.
Now accessible online, the tool allows hospitals to plug in numbers based on observations, including the number of COVID-19 patients, the average length of hospitalization and what’s known as the “doubling rate” — the time it takes for the number of cases to double in a community.
The tool then calculates how many gloves, gowns, masks and ventilators might be needed over the coming weeks or months — and shows whether a facility might be burdened beyond capacity.
Ofer Amram and Sterling McPherson, professors in WSU’s Elson S. Floyd College of Medicine, said their goal was to provide hospitals with a flexible tool that could be used for short-term preparations in a rapidly evolving outbreak scenario.
“Everything is a little bit calmer now, but a month ago it was just crazy. It was a lot of uncertainty,” Amram said. “People wanted to know projections and how we can handle that, so we worked really hard and really intensely to provide that.”
Amram and McPherson see the tool as complementary to other COVID-19 forecasting models, such as the widely cited projections from the University of Washington’s Institute for Health Metrics and Evaluation.
Amram said he was using the local model weeks ago to produce regular reports for MultiCare and the SRHD. McPherson said they also sent reports to the Yakima County Health District and a hospital chain in Western Washington.
One researcher from McPherson’s lab also introduced the tool to colleagues at the Federal University of Sao Paulo in Brazil, where some authorities are struggling to enforce social distancing measures.
