Even though the University of Georgia’s new number of COVID-19 cases could be an encouraging sign, the virus could still be rampant on campus. COVID-19 cases decreased by 1,000 the week of Sept. 7-13 — a drop of over 70% from 1,490 for the week of Aug. 31.
Mark Ebell, a professor of epidemiology at UGA, said in an email on Sept. 16 to The Red & Black that surveillance testing data showed the coronavirus is still impacting many students.
According to Ebell, the 7.61% positivity rate still shows that the virus is widespread on campus. “We need to continue to exercise caution on campus, and be much more careful off campus,” Ebell said. The testing positivity rate measures the percentage of tests that are positive, and UGA only reports a positivity rate for surveillance, or asymptomatic, testing. A high positivity rate suggests that there may be many cases that go undetected and that more testing may be needed.
The World Health Organization says countries should strive for a 5% positivity rate, but currently UGA has a positivity rate of 7.61%. The positivity rate from UGA’s surveillance testing could generally be lower because it focuses on strictly asymptomatic volunteers. Thus, a 5% goal may be too high. The university has said it strives for “a decreasing” positivity rate.
Other colleges in Georgia, such as the Georgia Institute of Technology, have been able to get outbreaks under control. In August, clusters of COVID-19 cases developed in Greek Life organizations at Georgia Tech. From Aug. 25-29, Georgia Tech’s five-day moving average number of daily cases from surveillance testing was 52.2, with a positivity rate of 2.6%. However, from Sept. 8-12, the five-day moving average of daily cases from surveillance testing was only 10.6, with a positivity rate of 0.7%.
That could certainly be happening at UGA too. Notably, the seven-day moving average of cases in Athens-Clarke County has fallen from 960 on Sept. 9 to 425 on Sept. 19, perhaps a sign that the situation has improved somewhat in the local community. However, data from wastewater samples has not shown the same decline, suggesting that the county may not have seen as much improvement as the drop in new cases would suggest.
UGA could still update last week’s case numbers as more test results are reported. The university frequently adds cases to previous weeks’ totals. UGA spokesperson Greg Trevor said in an email the numbers are revised when UGA reconciles its results with the Georgia Department of Public Health and when tests from one week have their results reported the following week.
Originally, UGA reported 1,417 cases for the week of Aug. 31. On Wednesday, UGA updated that figure to 1,490.
Notably, the positivity rate dropped by a much smaller margin than the number of cases reported. Although new cases dropped by over 70%, the positivity rate fell from 9.08% to 7.61%. If the positivity rate didn't improve as much as cases did, then there might not have been as much of an improvement as one might think.
This suggests that the drop in cases may be due in part to a drop in testing or self-reporting. Thus, the decline in new cases could be overstating the improvement in the situation.
Colin Smith, a professor of health management and policy at Georgia State University, said there could be outside reasons for the drop in cases reported from off-campus sources.
“Part of this may be delay in reporting, and part of this may be something that you’ve seen statewide already,” Smith said. “We’ve seen roughly a 25% drawback in testing statewide.”
It is impossible to know this for sure because UGA only provides testing information for its surveillance testing. The university likely does not have data on the number of tests taken off-campus.
Although a drop in the testing positivity rate could signal improvement, Smith believes it is likely too early to make conclusions about which direction the positivity rate is headed.
“We’re still dealing with numbers that are in flux,” Smith said. “ I would want to see if this keeps going down or if it starts going up. I’m not going to make a trend analysis really based on two or three points at this point because the program is still so new.”