A peak in the number of cases of the novel coronavirus in Colorado is likely still weeks to months away, coming in mid-September or later, according to projections released recently by a team from the University of Colorado.
Depending on how well the social distancing measures are working, they predict coronavirus deaths could number tens of thousands by the end of the year.
The number of deaths they predict changes dramatically depending on how much social interactions are reduced beyond the first set of restrictions implemented by policymakers from March 17 to March 25. They have not yet been able to predict the efficacy of stay-at-home orders put in place March 26.
Their projections contrast with a prominent coronavirus spread model produced by the Institute for Health Metrics and Evaluation at the University of Washington that predicted Colorado’s peak passed a few days ago. But the local researchers said they believe the IHME model, because it was based primarily on data from Wuhan, China, is not well suited for predictions about Colorado.
Over the past few days, state officials have released more of the expert research being relied upon for high-level decision-making about the state’s various social distancing measures in response to the coronavirus pandemic.
The newly released documents explain some of the factors that have gone into state public health officials’ decision-making, and updates predictions about how the virus will play out in Colorado. It also draws into focus the differences between the model produced by the University of Colorado School of Public Health and School of Medicine, and the models developed by other researchers.
“All models are wrong, but some are useful,” said Dr. Jonathan Samet, the lead researcher from the University of Colorado. The model they’ve developed will continue to be refined as more data about the cases in Colorado is collected and can be incorporated.
The University of Colorado model includes several assumptions that affect how the spread of the virus might unfold.
They assume the rate of infection, or the probability that someone with the virus will infect a susceptible person if they come into contact, is 41.3%. Dr. Kathryn Colborn, of the University of Colorado School of Medicine, said that for context, Measles, which is a much more infectious virus, has a rate of infection closer to 95%. Their model also assumes that symptomatic individuals are 2.3 times as infectious as asymptomatic cases.
The most recent research gives more detailed context to assertions made by state health department officials in recent weeks that the true number of cases is likely many times more than the number of cases counted in their own figures, which are a function of the number of tests being performed.
Because testing capacity was strained, many positive cases have surely not been counted officially. As of the end of March, the researchers believe 27.7% of cases were identified and tracked by the state, meaning the true number of cases was about 10,000 at the end of the month, even though the official figure was about 3,000.
Now the state has counted around 5,400 cases, and as of a few days ago, the researchers said, the true number of ill people is probably 17,000 or 18,000.
The researchers said they expect the portion of known cases that are counted will rise in coming weeks. The number of cases reported by state health officials is “just the tip of the iceberg,” Colborn said.
In the latest worst-case scenario modeled by the local researchers, where no social distancing measures were implemented or social distancing measures were totally ineffective, the number of cases would continue along a steep upslope, peaking in early May, which would correspond with 80,000 deaths by the end of the year — 73,000 deaths coming before June 1.
To get the spread to decline, without going through that scenario, the researchers predict social interactions must be reduced by 60%-70%.
The researchers highlighted in their most recent report a change in the curve starting the last week of March. They say the shift suggests social distancing measures may have already reduced social interactions by 45%, preventing around 1,200 cases prior to April 1 and thousands more by now.
The researchers pointed to vehicle traffic data as another indication that social interactions have been significantly reduced, even though the reduction in vehicle traffic doesn’t precisely correspond to social interaction reduction.
Even with a 60% reduction in social interactions drastically reducing the death count, modeling the state is relying on predicts at least 43,000 deaths by the end of the year, with a slower climb, peaking in mid-September. If the social distancing measures achieve a 70% or 80% reduction in social interactions, the number of predicted deaths declines dramatically, to 1,400 deaths.
The statewide “stay-at-home” order, the researchers believe, will begin to have an effect on the case numbers in the coming days, “perhaps by the end of this week,” Dr. Samet said.
State epidemiologist Dr. Rachel Herlihy said the coming week will provide new information that should give new and narrowed insights into how social distancing has worked and how the rest of the coronavirus’ spread will play out.