UW model says Washington could start opening May 18, but don’t mark the calendar yet
It was predicting the number of total cases and deaths from the coronavirus, but the latest model from the Institute for Health Metrics and Evaluation has something new to offer – a possible date that states across the country can start safely easing restrictions.
For Washington state, that sweet spot is May 18, according to the model.
The IHME model has been one of the go-to models for President Trump and Governor Inlsee as they determine next steps, despite recent criticisms from epidemiologists about the model’s accuracy and use by policymakers.
In this latest model, released Friday, IHME’s Dr. Christopher Murray tried to project when the rate of community infection will be so low, that even with testing capacity and contact tracing available at that time states could start to ease restrictions with a reduced risk of resurgence.
The dates range from the week of May 4 to the week of June 8 or later, and are determined by cases and deaths, when a state enacted social distancing measures, and how stringent they were, among other things.
For Washington state, the model suggests restrictions can start being lifted on the earlier end of that range.
“The results suggest that some aspects of social distancing could be considered as early as the week of May 18,” Murray said.
While May 18 might sound reasonable, there are still concerns.
“They are likely to change for two reasons,” Murray said adding one reason will be future information added to the model about what happens with cases and deaths after an individual state peaks. At issue is whether they drop quickly or slowly decline, which makes a difference on the timeline.
The second factor that could lead to fluctuation in the model is how well states ramp up testing and contact tracing, and isolation capacity.
“Greater capacity means that it might be reasonable to relax sooner and handle a bigger caseload in the community,” Murray explained.
He stressed that individual states will also need to consider the risk of importation of new cases from other states, especially neighboring ones that may have higher case counts than they do.
“If I were a governor of a state, I certainly would not make a decision just based on our model,” Murray said.
“I would recommend both looking at these types of models, but also looking at a range of indicators,” Murray added, such as capacity of the public health service, available surveillance and any random sample testing that might be available in a community.
Most recently, Murray’s model has been highly criticized because it fluctuates so much. Originally, U.S. deaths were predicted to hit six figures. The model now suggests deaths will be about 60,000.
For Washington state, the model originally put deaths around 1,400, but that number is now predicted to be about half at 694.
Murray has stressed from the beginning the model fluctuates based on new data being added all the time, and simply because of the type of model it is – statistical not epidemiological.
He also noted fluctuations in case numbers and deaths have fallen between the ranges included with each update to the model, wide ranges of as much more than 125,000 in some cases.
The key point to remember while looking at the IHME model – including all local, state, and federal decision makers – is that it is statistical based on ever-changing data. By nature, the model will change. Most models over the past century are epidemiological. They look at the science and actually track how the virus behaves, which is not something the IHME model even considers.
Several of those epidemiological models, including one released from Bellevue-based Institute for Disease Modeling, says the state’s social distancing measures are working as intended, but still must continue to remain in place, as Governor Inslee recently pointed out.