Dr. Daniel Carey is Virginia's secretary of health. (Ned Oliver/ Virginia Mercury)

Virginia’s social distancing measures are slowing the spread of COVID-19, the disease caused by a new coronavirus, according to a new model from scientists at the University of Virginia.

But the model, presented Monday by state health officials and researchers from UVA’s Biocomplexity Institute, also projects that case numbers will eventually exceed the surge capacity at Virginia hospitals without further mitigation strategies. 

The models are complex and the predictions far from certain, Virginia Health Secretary Dr. Daniel Carey said during Monday’s briefing. But the takeaways for state officials are straightforward. 

One, social distancing is vital to flattening the curve of infections and reducing the strain on the state’s hospitals and health care systems, he said. And two, Virginia still needs to develop ways to ease restrictions in the future without triggering future surges. 

“Our actions going forward will affect our outcomes,” Carey said. “We’re in this for the long haul, and Virginians need to keep working with us.”

The long-anticipated modeling from UVA paints a different picture than other projections from research groups across the country, including a widely cited model from the Institute for Health Metrics and Evaluation and the University of Washington. 

Unlike other models, UVA’s projections include county-level case counts and simulate how individual communities interact among themselves and other areas of the state. Researchers took limited testing into account, assuming that one confirmed COVID-19 case equates to about seven infections, said Bryan Lewis, a computational epidemiologist at the Biocomplexity Institute.

There are a few other assumptions built into the model, including that one infection will lead to between 2.1 and 2.3 more. Lewis said that researchers also assumed an infectious period between 3.3 to 5 days.

As scientists learn more about COVID-19, it’s become increasingly clear that the disease can incubate up to 14 days before the onset of symptoms and that it can spread before symptoms begin. But the UVA model assumes that most patients will fully isolate themselves before then, shortening the infectious period, Lewis added.

The current model could change based on policy decisions and other unknown factors, such as how the weather affects transmission. As of Monday, though, there were five different scenarios that differed based on transmission speed, the effectiveness of current policy measures and the date when social distancing measures were lifted.

The best case assumes that social distancing measures stay in place until June 10. It also projects that the restrictions are pausing coronavirus transmission in Virginia, which would stabilize day-by-day case increases.

Under that scenario, Virginia would see a later, lower peak between mid-July to late August with around 15,000 confirmed cases. The surge would exceed hospital capacity in some regions around the same time, though researchers pointed out that continued social distancing could postpone that by one to two and a half months.

That’s far fewer than the worst-case projection — one that assumes transmission continues unchecked. In that case, Virginia’s peak would arrive between late April or mid-May with more than 35,000 cases. Under the model, the surge would exceed hospital capacity around the same period.

That scenario is improbable based on the observed impact of Gov. Ralph Northam’s current policies, which largely started on March 15 when he banned large gatherings and continued through March 30 when he issued a stay-at-home order.

Since then, known COVID-19 cases in Virginia have risen significantly, from 45 on March 15 to 5,747 as of Monday, less than a month later. And under a more moderate scenario — one where transmissions slow, but don’t stop, and the stay-at-home order extends through June 10 — the state would see a peak sometime between early May and mid-June with between 15,000 and 20,000 cases. At least some regions would see caseloads exceed capacity around the same time. 

But experts also cautioned that the peak number of cases in Virginia could be lower than even best-case projections if social distancing measures continued past June 10 or other policy decisions came into play. Lewis said the model assumes that business, travel and social interactions would return to roughly half their normal rate after restrictions were lifted, which would still speed up the spread.

“We really think that the way we’ve shown the peaks are never going to happen,” added Madhav Marathe, director of the Network Systems Science and Advanced Computing Division at the Biocomplexity Institute. “People always adapt to the ground reality. So, the peak should be viewed as a mathematical possibility, but our belief is that it’s going to guide the interventions that are put in place.”

What’s not being modeled is how the state will eventually decide when to lift restrictions — and how to do it in a way that prevents another outbreak. There are nationwide concerns over the availability of testing and how states can determine which residents have already been infected. Virginia officials recently announced a contract with McKinsey in an effort to boost testing capacity.

It’s up to the state to develop new resources by the time Virginia reaches its peak this summer, Carey said. To reopen businesses and travel, officials need to know how many residents have been infected with COVID-19 and how many are still at risk of catching or transmitting disease.

He specifically mentioned antibody tests, which can determine which patients have already been exposed to the disease, along with widespread screening and recently proposed cell phone apps that could trace the spread of disease

“We definitely need to develop a different toolkit besides opening everything up or opening it halfway,” Carey said. “There are a variety of tools — some of which, frankly, have not been developed yet — that we need to develop between June and July and August to make sure we do this in a nuanced way.”