Created by the University’s Institute for Health Metrics and Evaluation, the model encompasses mortality data revealed by Johns Hopkins University, as well as pieces of information from WHO (World Health Organization) and governments from around the globe. While talking to CNN, the institute's director Chris Murray said the increased estimation is a result of the states beginning to include nursing home deaths under the COVID-19 case category.
The recently updated model shows an estimate of when social distancing restrictions can be lifted in each state, noting that the timetable is based on the measures taken to restrict the spread, isolation, contact tracing, and restricting large crowds. States that are in a position to consider relaxing from May 4 to May 10, according to the model, include Vermont, North Carolina, West Virginia, Montana, Hawaii, and Alaska.
The model also shows states that should consider waiting until Jun. 8 before reopening businesses. These states include Florida and Georgia, both have been strongly criticized for lifting restrictions too quickly. If people ignore social distancing and go back to interacting normally with other people, the risk of transmission will rise drastically, Murray warned.
Former institute employee, epidemiologist Ruth Etzioni told Stat News that the model is a highly unreliable predictive tool, given that it keeps changing. She cited the model's Apr. 17 update. Etzioni noted that if it is used for policy decisions, there are chances of its results been interpreted wrongly, adding its a travesty unfolding before our eyes.
It is imperative for model outputs to adjust to new data and when while dealing with nonlinear systems, even small changes can trigger a massive effect, Nate Silver of FiveThirtyEight tweeted responding to Etzioni’s comments. He deemed the model's estimated death toll reasonable while claiming it is an estimated three- to fivefold decrease in hospitalizations wasn't great since the revision was large.
Forecasts of deaths will come in handy for public health decision making as they provide an estimate of upcoming impact, according to the CDC (Centers for Disease Control) website. The CDC says it considers data from various organizations including the University of Washington model to improve its efforts, but the White House considers the Washington model along with another model created by Imperial College in London as revealed in a Mar. 31 press briefing.
Imperial College suggests about 2.2 million people could die in the United States if mitigation efforts such as social distancing were ignored. At the time, the University of Washington model hinted 38,000 to 162,000 deaths would occur through summer.