‘The data are imperfect:’ Searching for more precise ways to track coronavirus spread


RALEIGH, N.C. (WNCN) – With the number of coronavirus cases expected to keep rising in the near future, doctors and health officials want to more precisely measure and understand its spread.

Those numbers are climbing both because the virus is exceptionally contagious, and because more people are being tested.

The number of confirmed cases of COVID-19 in North Carolina from March 18-30.

Dr. Gregory C. Gray — an epidemiologist, Duke professor and expert on emerging infectious diseases — called the data “imperfect.”

“We know the count has increased as testing availability increases,” Gray said. “So we are only catching the people that are, No. 1, sick enough to seek care or not too sick but suspected and then able to get the testing. There’s a whole lot of circulation of the virus going on under the radar and under the clinical radar, so to speak, and how much of that, we don’t know. There have been some modeling estimates, but it’s not based on a lot of facts.”

Just in North Carolina, the Department of Health and Human Services reported 1,307 positive cases in the state – more than twice as many as there were four days earlier — and six deaths following 20,864 tests, according to the agency’s update Monday.

The percentage of coronavirus tests in North Carolina that have come back positive has steadily risen during the past two weeks.

That yielded an overall positive test rate of just over 6.2 percent in the state. That rate was 3.4 percent on March 18. But with a state population of 10.3 million, according to the U.S. Census Bureau’s most recent estimate from 2018, that means just 0.2 percent of the population has been tested.

“We have to acknowledge we’re just at the beginning, and every indication is, this is really ramping up now, and we’re in what we would call the acceleration phase,” State Epidemiologist Zack Moore said Monday. “We certainly have not peaked. … You can just look across the country and look across the world at the trends, and we’re still on the way up.”

Those numbers can be volatile. On one day last week, only 2.3 percent of the 2,043 recorded tests confirmed coronavirus infections. A day later, 6.2 percent of the 1,162 recorded tests were positives.

Dr. Elizabeth Cuervo Tilson, the state’s health director and chief medical officer, says that while private labs are required to report positive tests, that isn’t always the case with negative results. That has led to some imbalanced numbers. One daily update included increases of 101 confirmed cases – and just 64 total tests.

So, it’s hard to draw many valid conclusions from those statistics.

“Laboratory case counts are never the whole picture,” Moore said.

Tilson says the best way to understand the statewide spread “is not to try to test everybody” but to apply the evidence-based surveillance strategy used by the state to track the spread of the flu “and be able to apply that to COVID-19. That will give us much more evidence-based, science-driven data on the spread of the disease.”

Gray – who leads the Duke One Health Research and Training Network, which conducts research and training in 14 countries – said Duke would like to conduct a limited epidemiological study of 250 hospitalized patients and 1,250 people who qualify as their close contacts.

That study would draw blood from the patients at the start of the study and after 14 and 28 days, with blood taken from those close contacts at the beginning and after three weeks.

“We’ll get a quick look at immediate secondary transmission, how much of it is subclinical,” Gray said. “But it won’t be the quality of the data that you would get in a larger study where we could stratify, ‘Well, if you’re in this age group, 60-70, this is your risk,’ and those sorts of things are important in getting answers to epidemiological questions.”

The most detailed answers would come from a comprehensive, population-based study that would involve analyzing blood specimens from a large group of people and follow that with repeated blood draws that would be checked for evidence of infection and antibodies, he said.

An obvious starting point, he said, would be to piggyback off an already-underway group study – also known as a cohort study – focused on people with respiratory diseases.

“They’re already set up,” he said. “They’ve got the people engaged. They’ve got (blood) on them for the last year, and it would be relatively easy for them to adapt their cohort studies to meet the new requirements.”

As a general rule, studies like those can be costly: One involving monthly blood draws of 2,000 people for a year would cost roughly $2 million, he said, adding that they often require a major sponsor or governmental support because “you can’t just go do that with existing funding.”

Gray said the U.S. Centers for Disease Control and Prevention have sought proposals for such studies.

“These studies are how we make decisions regarding chronic care,” Gray said. “We need that sort of comprehensive study for a population here in the United States right now to better understand how the coronavirus is spreading.”

Copyright 2022 Nexstar Media Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.

Trending Stories

Click here for full list of trending stories