R. Kelly Jordan
As the COVID-19 crisis wears on, many know that doctors and other healthcare providers are working to better understand the disease and how to best fight it. What many don’t see is another group of professionals in the same fight, whose work is just as important but from another angle – the number-crunchers.
Health policy experts, economists, statisticians and others often create “modeling projections” as tools to provide a sort of educated guess as to a particular health-related outcome, and shape health policy decisions.
Hospitals are staffed and equipped, generally, in proportion to the number of patients they can expect to treat over a given period of time.
In an area such as Hardin County, HMC can expect to see a relatively stable number of medical issues such as heart attacks, strokes, broken bones, flu or pneumonia, and there are models that bear that out. So, when an immediate threat like COVID-19 appears, models are used to estimate the impact of the threat on the hospital, and help guide decisions on how to prepare.
Considering the size and scope of the COVID-19 pandemic, numerous models have been created around the world.
There are models estimating how fast the disease will spread in a given population or area, how many people will die from the disease, how much equipment or drugs will be needed, or what effects social distancing and stay-at-home orders will have on the spread of the disease.
At a recent Hardin Medical Center Board meeting, Chief Operating Officer Rebecca Wright, who assists HMC with strategic planning along with other responsibilities, told the board about modeling projections HMC has used to prepare for COVID-19 and care for all patients during the crisis.
In addition to in-house modeling done by HMC employees using a “surge demand calculator” called SG2, Wright said HMC has used a number of different models to try to predict and prepare what might happen in Hardin County, including the “Vanderbilt Model,” which was created at Vanderbilt University and is Tennessee-specific.
Dire early predictions
Wright noted that early on, in February and March, and prior to the Vanderbilt Model, most modeling predicted that an average of about 40% of an area population would quickly become infected with COVID-19 after it began to spread in that area.
“What we saw happening across the world was alarming, because the healthcare system couldn’t keep up with that 40% infection rate they were predicting,” Wright said.
So, Wright said, HMC looked at what would happen if Hardin County were to reach the 40% infection rate.
“Some of these numbers were quite alarming at the beginning. If you take the population of Hardin County and say 40% become infected, you’re looking at over 10,000 residents.”
Wright said based on studies from around the world, in areas that were at the height of their own COVID-19 crisis, 15% of those infected were expected to need hospitalization – meaning over 1,500 hospitalizations for Hardin County alone, based on 10,000 infections in the county.
Worse, that would potentially happen over a period of 60 to 75 days, amounting to more than 200 hospitalizations per day, just at HMC. Currently, she noted, HMC sees about 1,400 hospitalizations a year; the deluge would easily overwhelm the hospital.
Complicating matters, Wright pointed out, is that there would be no place to transfer patients HMC couldn’t care for, because the hospitals in the entire region and state would be in the same overwhelmed condition. HMC would be on its own.
As Tennesseans adopted social distancing practices around the middle of March, however, Wright said the models began to be downgraded. She said also that in the U.S., the experience has proved that a smaller percentage of infected are needing hospitalization, leading to even less pressure predicted for hospitals – good news, but still overwhelming.
The models now showed, around mid-March, an average of 10 “normal” and 55 or so COVID-19 inpatients at HMC at any one time, and still more than HMC can handle.
Then, in the first week of April, just as Lee issued the stay-at-home order, the Vanderbilt Model was published. This was the first widely-recognized model created specifically for the state, regions, cities and communities of Tennessee, using health data specific to Tennessee.
The Vanderbilt Model offers three predictive models based on social distancing efforts; the “optimistic scenario,” meaning extra social distancing efforts beyond those of early April; the “status quo scenario,” if efforts remained the same; and the “lift scenario,” if social distancing measures were relaxed.
Under the optimistic scenario, Tennessee would achieve a halt in the rise in COVID-19 cases by mid-May, and hospitals would not be overwhelmed. With status quo, the Vanderbilt Model predicted a peak in hospitalization in mid-June, and possible stress on the hospital system in the state. The lift scenario predicts that hospitals will be overwhelmed if done too soon, and without robust testing and contact tracing policies in place. The lift scenario predicts that over 50,000 Tennesseans will be sick enough to require hospitalization.
Wright told the board that under status quo, Hardin County and HMC were predicted by the Vanderbilt Model to need to hospitalize about 20 patients a day – much lower than the earlier predicted 55, and very much lower than the earliest predicted 200.
She pointed out that the reason social distancing measures work is because they slow the spread of the disease among the community, not completely stop people from getting the disease.
“That’s what ‘flattening the curve’ means. People are still getting infected, however they’re getting infected over a longer time period, which helps healthcare providers better care for them because we have fewer patients at any one time.; We have fewer patients needing care per day over the long term,” Wright said.
She also noted that slowing the spread gives healthcare providers time to assess possible new treatment discoveries.
Hardin County sees lower incidence of infections
To date, Hardin County has seen a much lower incidence of COVID-19 infections than anyone predicted, and no local hospitalizations. Wright, and the Vanderbilt Model, attribute that to social distancing measures by Hardin countians and Lee’s stay-at-home order.
Local uptick expected
That order expired Thursday, and while Lee is publicly urging all Tennesseans to continue to practice social distancing measures, Wright says HMC predicts an uptick in COVID-19 infections.
The Vanderbilt Model has predicted that if infections begin to rise enough, another round of restrictive social distancing and stay-at-home orders may have to be enacted, either localized to a specific area or statewide. In a comparison of widely recognized models from around the world included in the Vanderbilt Model paper, virtually every other major model predicts that a series of social “stops and starts” will be required in almost every area, almost regardless of the level of social distancing.
Using variations in the rate of spread of COVID-19, denoted by “R,” the Vanderbilt Model predicts, if the rate rises, how many days it will be until the economy may need to be shut down again.
The rate of spread for COVID-19 in Tennessee has been above 5, meaning each infected person is expected to infect five others. Currently, the rate of spread is estimated to be very close to 1. The rate of spread for flu averages about 1.5, and for a communicable disease to disappear in a community, its rate of spread, or R, must fall below 1 and stay there.
The most current projections of the Vanderbilt Model predict that, after the stay-at-home order expires on May 1, if the R in Tennessee stays at or below 1, hospitals statewide will be able to handle those who are sick and the economy should remain open.
If the R rises very slightly to 1.1, the economy will likely remain open for 129 days, and will need to be shut down again around the end of August or beginning of September.
If the R rises to just 1.2, the economy will need to be shut down again in just 74 days, meaning mid-to-late July.
However, if, after the stay-at-home order is lifted the R rises to just 1.5 – about the same rate of spread as the flu – the Vanderbilt Model predicts all of the social distancing gains so far will be lost and the economy will need to be shut down again in just 48 days.
Depending on how seriously Tennesseans take social distancing over the next weeks and months, we could be back to square one before the end of June, or we could be on track to beat this bug.