IHME Adds More ‘One Size Fits All’ Social Distancing Assumptions to Updated Coronavirus Projections

TOPSHOT - Two women observe social distancing measures as they speak to each other from adjacent park benches amidst the novel coronavirus COVID-19 pandemic, in the centre of York, northern England on March 19, 2020. (Photo by OLI SCARFF / AFP) (Photo by OLI SCARFF/AFP via Getty Images)
OLI SCARFF/AFP via Getty Images

The Institute of Health Metrics and Evaluation (IHME) released an updated version of its coronavirus model on Friday that added two more “one size fits all” assumptions about the effectiveness of government imposed social distancing restrictions on all 50 states.

The IHME model now includes “six categories of social distancing policies,” four which have been included since the March 30 update (stay-at-home orders, educational facilities closed, non-essential services closed, and travel severely limited), and two of which were added in the model released on Saturday (mass gathering restrictions and initial business closure).

As Breitbart News reported, the IHME model’s credibility has been dramatically diminished in recent weeks by its extraordinarily high overprojections of the need for regular hospital beds and ICU hospital beds.

Indeed, the alarming projection in the initial publication of the IHME model, as published on March 26 that, “In addition to a large number of deaths from COVID-19, the epidemic in the US will place a load well beyond the current capacity of hospitals to manage, especially for ICU care,” (emphasis added) was far off the mark.

The model’s original forecast relied on “data on confirmed COVID-19 deaths by day from WHO websites and local and national governments; data on hospital capacity and utilization for US states; and observed COVID-19 utilization data from select locations to develop a statistical model forecasting deaths and hospital utilization against capacity by state for the US over the next 4 months.”

The subsequent IHME model forecasts do not appear to rely on other factors such as population density and mass transit usage, which at least one study suggests are factors in the spread of COVID-19, resulting in “one size fits all” social distancing assumptions applied to states where population density and mass transit usage are dramatically different, such as New York on one extreme and Wyoming on the other extreme.

The latest version of the model is designed to influence the dates governors choose to reopen their states by “expanding categories of social distancing policies and accounting for mobility,” according to the update provided by the IHME with the release on Friday of new projections:

Social distancing policies, which can range from restrictions on large gatherings to strict stay-at-home orders and closure of all non-essential businesses, have been used as a mechanism to substantially reduce the spread – and thus the immediate toll – of COVID-19. We are now entering the phase of the epidemic when government officials are considering when certain types of distancing policies may be eased. With today’s release, we provide initial estimates that can serve as an input to such considerations in the US.

These estimates assume that when social distancing policies will be eased, such actions will occur in conjunction with public health containment strategies. Such measures include widespread testing, contact tracing, and isolation of new cases to minimize the risk of resurgence while maintaining at least some social distancing policies to reduce the risk of large-scale transmission (e.g., bans on mass gatherings).

The model now not only projects deaths and hospital capacity requirements, it also presumes to determine the date in each state when “relaxing social distancing may be possible with containment strategies that include testing, contact tracing, isolation, and limiting gathering size.”

The model’s coronavirus death projections, as updated on Friday, declined to 60,308 in the United States by August 4, down from a death projection of 68,841 in the April 13 update, which was down further from a death projection of 93,765 in the April 1 update. As of April 21, slightly more than 40,000 deaths in the United States have been attributed to coronavirus, COVID-19.

On Monday, several governors announced they will begin to open their states up on May 1,  much earlier than the developers of the IHME model assert “relaxing social distancing may be possible,” as Breitbart News reported:

A trio of Southern states announced separately Monday that they plan on loosening stay-at-home orders implemented as part of an effort combat the spread of the Chinese coronavirus.

Georgia Gov. Brian Kemp (R) detailed plans to allow some businesses, including gyms, salons, and related indoor facilities, to reopen by Friday. Workers and patrons must social distance and meet several other safety guidelines. Restaurants will be allowed to reopen on April 27, while bars will remain closed. . .

Georgia’s shelter-in-place order will stay in place until April 30.

In Tennessee, Gov. Bill Lee (R) also announced he will not extend his state’s stay-at-home order past April 30 and that several types of businesses could begin opening their doors by week’s end. . .

Additionally, South Carolina Gov. Henry McMaster (R) said some retail business will be allowed to reopen which had been designated as “non-essential.”

“Our measured, deliberate approach has been the right one we believe,” explained McMaster. “Our goal was to cause the most damage possible to the virus, while doing the least possible damage … to our businesses. South Carolina’s business is business.”

Barbershops, salons, and gyms will remain closed until further notice, the governor said.

The updated IHME model released on Friday, however, projects Georgia should not begin opening up until June 15. The model projects a safe date to begin opening up South Carolina at June 1, with Tennessee’s date at May 25.

According to Friday’s IHME update:

Since our last release, our team has expanded the types of social distancing policies included in our COVID-19 death model. Further, our social distancing covariate now accounts for the effect of different types of measures on population-level movement patterns over time, drawing from mobility data collected from cell phone and online platforms. In combination, these updates now more accurately reflect the effect of social distancing policies enacted – and importantly, how people are changing their behaviors in response to these measures.

Here is a summary of how those categories are defined:

Stay-at-home order: “Individuals are ordered to stay at home and interactions between individuals from different households prohibited.”

Educational facilities closed: “Education has transitioned to remote-learning options.”

Non-essential services closed: “All stated non-essential services are clearly defined and ordered to cease operations.”

Travel severely limited: “Municipal-level travel restrictions are in place and movement is strictly limited to essential travel. Public transit options are reduced and international borders are closed.”

Mass gathering restrictions: “Initial implementation of restrictions on people gathering are mandated.” This is a social distancing policy newly included in the IHME COVID-19 estimation effort as of April 17. Recommendations wherein individuals are urged to not congregate en masse without a formal mandate are not included.

Initial business closure: “The first phase of non-essential businesses are ordered to cease on-premise operations.” This is a social distancing policy newly included in the IHME COVID-19 estimation effort as of April 17. Examples include the mandated closure of specific groups of business types (e.g., recreational and fitness facilities, entertainment venues, personal care services) without calling for a more sweeping closure of non-essential businesses.

The IHME model appears to have not included certain “municipal level travel restrictions” put in place in New York City that MIT Professor Jeffrey E. Harris wrote in a recent National Bureau of Economic Research working paper had the effect of likely increasing the spread of the coronavirus, as Breitbart News reported:

A National Bureau of Economic Research (NBER) working paper by MIT Professor Jeffrey E. Harris concludes that the New York City subway system “seeded the massive coronavirus epidemic” in the city.

Even more significantly, the Metropolitan Transit Authority’s (MTA) decision to reduce subway service exacerbated the problem because the more crowded subway cars “most likely accelerated the spread of coronavirus.”

Metropolitan New York City remains the epicenter of the coronavirus pandemic in the United States. More than half of the more than 40,000 coronvirus deaths in the United States as of April 19 have taken place in two states: New York and New Jersey. The bulk of the deaths in those states have been in the Metropolitan New York City area, which is not only the most densely populated region of the country, it also has by far the heaviest daily usage of mass transit, particularly subways.

“New York City’s multitentacled subway system was a major disseminator – if not the principal transmission vehicle – of coronavirus infection during the initial takeoff of the massive epidemic that became evident throughout the city during March 2020,” Harris, who has a Ph.D. in Economics and an M.D. and is both a practicing physician and a professor of economics at MIT, wrote in the abstract of the NBER working paper published on April 13.

The current version of the IHME model projects that New York State will be safe to begin opening up on June 1, but also states that travel limitations are not in effect even though the New York’s Metropolitan Transit Authority cut the number of subway cars in operation by an estimated 25 percent in March.

The next update of the IHME model, likely to be released this week, will factor in the new social distancing rules as announced by each governor in their projections. States such as Tennessee, South Carolina, and Georgia, which have announced those governmental restrictions will be relaxed as of May 1, are likely to see an increase in projected deaths and hospital utilization in subsequent IHME model updates, setting the stage for even greater political controversy.



Please let us know if you're having issues with commenting.