Projection shows COVID-19 cases could nearly double by inauguration
The number of COVID-19 cases in the United States could nearly double — or worse — before the inauguration of President-elect Joseph R. Biden on Jan. 20, according to a study published in Scientific Reports.
“In the early days of the pandemic, we were frustrated by the poor forecasts that we saw,” Raphael Thomadsen, PhD, a professor of marketing at the Washington University in St. Louis Olin Business School, told Healio. “These forecasts either had a tendency to predict a short-term catastrophic outcome that wasn’t happening or to predict that the pandemic would end by the end of the summer, which was clearly implausible as well.”
Thomadsen and colleagues developed their own model, which accurately predicted the rate of COVID-19 growth over the summer.
“Underlying these bad forecasts was an assumption in other models that growth rates had to be proportional to the number of people who got sick,” Thomadsen said. “Thus, the functional form of the models forced the model to say that growth would be either exponential or declining.”
Instead, Thomadsen and colleagues “observed increasing growth rates, but the growth rate was more linear,” he said.
20 million cases
Their projection was a modified version of a standard susceptible-infected-recovered model. It combined COVID-19 case data with social distancing figures based on cellphone GPS location data. According to a press release, the model also incorporated the percentage of a county’s population without a confirmed cases of COVID-19, the number of individuals who have been infected each day and the county’s population density.
Different from traditional models, which assume that people interact randomly with other people, Thomadsen said their model notes that friends tend to interact with a common group of friends, and people work with a common group of coworkers.
“This type of interaction was not incorporated into the other prominent forecasts, so we wanted to incorporate that,” Thomadsen said. “Once we allowed the model to condition for these network effects, we found that the model fit a lot better.”
Based on the numbers of U.S. cases reported by the Johns Hopkins COVID-19 dashboard at the time of their paper 11.4 million Thomadsen and colleagues’ model projected a rise to around 20 million cases by the end of January. This is based on current social distancing measures, which reflect a 60% return to normalcy. When adjusting the model to a higher or lower rate of return to normalcy, the researchers explained that even small increases in social distancing could have a significant impact. For example, a 50% return to normalcy, similar to what the country experienced in August, could result in around 5 million fewer cases by the end of January.
Unfortunately, the projections could potentially be even more grim. According to Thomadsen, the model was not built to include potential effects from holiday travel and gatherings. He explained that the model does predict the rate of spread as a function of the level of social distancing that occurs, allowing researchers to forecast that if a holiday causes a decline in social distancing, then it should lead to more spread and a wave of further cases.
He said the estimation of cases doubling by the time Biden takes office does not account for these gatherings.
“This means that our forecasts are likely to be on the conservative side,” he said.
On the other hand, Thomadsen said they cannot yet anticipate how many governors will add restrictions on movement. If enough governors act, he said it is possible that contact will be significantly reduced and then the spread of COVID-19 could also be greatly reduced.
A model developed at the University of Washington showed a similarly grim forecast, projecting around 375,000 deaths by mid-January, equating to 2,500 deaths per day in the U.S. Looking beyond January, the projection estimates roughly 471,000 deaths by March 1.
According to Ali Mokdad, PhD, chief strategy officer of population health at the University of Washington’s Institute for Health Metrics and Evaluation, this will be the trajectory if states take steps to control the spread, such as closing nonessential businesses, limiting gatherings and adhering to mask mandates. If states take no extra measures, deaths are projected to reach 417,000 by mid-January, whereas if 95% of the population always wore a mask when leaving home, that number could be reduced to 345,000 deaths by mid-January and 406,000 by March.
“These models are a crucial tool for policymakers to plan for the surge and pandemic,” Mokdad told Healio. “The models also help inform the public of the situation and stress the importance of mask wearing and other steps that individuals and policymakers can take to reduce risk.”
He said a plan must be put in place for a “winter surge” because the rising numbers of cases mean that there will be pressure on hospital resources in the coming months.
“The public must do their part to reduce spread and not overwhelm our hospitals,” he said.
The main message, Thomadsen said, “is that the next 2 1/2 months of COVID might be as bad as the previous 7 months combined.”
“There is still time to act,” he said, “but if we, as a country, do not have the will to pull back on our activities, we will see COVID cases double in a very short time period.”
Johns Hopkins. Coronavirus Resource Center. https://coronavirus.jhu.edu/us-map. Accessed on Nov. 24, 2020.
The Source. COVID-19 cases could nearly double before Biden takes office. https://source.wustl.edu/2020/11/covid-19-cases-could-nearly-double-before-biden-takes-office/. Accessed on Nov. 24, 2020.
The Source. Study: Researchers forecast stable, slightly declining levels of COVID-19 entering fall. https://source.wustl.edu/2020/06/study-researchers-forecast-stable-slightly-declining-levels-of-covid-19-entering-fall/. Accessed on Nov. 24, 2020.