Even with increased testing, predicting the spread of coronavirus remains a challenge
The growth in coronavirus cases in North Carolina has far outpaced an early forecast, but it’s hard to know how much of the growth is attributable to increased testing and how much to increased infection rates.
The 272 cases reported last Saturday surpassed the “very preliminary” projection for 110 coronavirus cases made by UNC-Chapel Hill epidemiologists, whose early model did not take into account expanded testing or state orders enforcing social distancing.
Kimberly Powers, an associate professor of epidemiology at UNC-Chapel Hill, said Monday that a more refined model is in the works.
“Taking into careful account how these factors interact is certainly our charge,” she said. The preliminary model forecast 1,000 cases in the state by the end of this week and 4,000 by April 2, The News & Observer has reported.
The UNC model is not the only information hospitals have.
Last week, Dr. Joseph Rogers, chief medical officer for Duke University Health System, said the university was developing its own forecasts based on data from Wuhan, China, Washington state, and the Lombardy region of Italy, The News & Observer has reported.
New infections and expanded testing are contributing to the increase in reported cases, said Dr. Christopher Ohl, an infectious disease expert with Wake Forest Baptist Health, based in Winston-Salem.
“The increase in cases in North Carolina is due to a rise in new infections from day to day, but also, due to an increase in testing, particularly in some regions where testing was more scarce one to two weeks ago,” Ohl said in an email.
“The Charlotte area in particular seems to be having a bit of a faster increase in true new infections. It will take a couple more days to get a better idea. Our state, however, at this time, is not seeing nearly the rise in infections that New York and San Francisco are.”
The state reporting on testing has changed over time. In the early days, only tests performed by the state laboratory were reported. Information on tests performed by private labs and hospitals came on later.
Figuring out how much expanded testing has contributed to the increase in COVID-19 cases is difficult, said mathematical biologist Alun Lloyd, Drexel Professor of Mathematics at NC State University.
“In order to be able to properly interpret the number of cases, you really need to know the number of tests being performed,” he said in an email.
“Another issue (more related to the amount of spread rather than interpreting the number of cases) is the fact that some infections are asymptomatic,” he wrote. “If someone doesn’t show symptoms, they won’t show up in the cases and would be unlikely to be tested. But they might still be causing infections.”
The state is going to rely less on testing and more on other information collection techniques as it tracks the spread of COVID-19.
People with mild symptoms can recover at home, and may not need to go to a medical office, Dr. Mandy Cohen, head of the NC Department of Health and Human Services, said at a news conference Monday.
“You can call the doctor to see if you need to be seen,” she said. “Most people will have mild illness and recover at home.”
Epidemiologists will use “surveillance techniques to make sure we fully understand the virus as we go forward,” she said.
She did not describe how DHHS would collect the information specifically for the coronavirus, but its communicable disease branch is responsible for tracking diseases and has a flu surveillance program, according to its website.
At a news briefing Friday, Dr. Betsy Tilson, the state’s medical director, said testing won’t be as important as the state moves to mitigation.
“We’re putting into place other surveillance strategies in order to understand the spread of the disease,” Tilson said.