“By not reporting antigen tests in the daily charts and maps, it distorts the picture and it distorts it at the local level where people make decisions based on this data,” said Dr. Harry J. Heiman, a clinical associate professor at the Georgia State University School of Public Health.
How big is the distortion?
An Atlanta Journal-Constitution analysis of DPH data through Wednesday shows the state’s two-week per capita case rate would be about one-third higher if antigen positives were included in the calculations with the gold standard PCR tests. On Wednesday, DPH reported a rate of 313 infections per 100,000 people, or three times the level DPH considers as substantial spread. Including antigen tests, the AJC found the two-week case rate that would grow to 412 infections per 100,000, or four times the substantial spread level.
The seven-day rolling average of reported cases would be 3,362 cases per day through Wednesday if antigen tests were included, or about a third more than PCR tests alone. That rolling average would be about 10% below Georgia’s peak reported daily case average level in July (though test positivity now for PCR tests is lower, suggesting Georgia is capturing more cases than in the summer).
In 60 of Georgia’s 159 counties, the two-week case rate would climb by 50% or more, including in 16 counties where the rate would be double or more than what is published daily on DPH’s website, according to the most recent county-level antigen data examined by the AJC.
In Morgan County, 60 miles east of downtown Atlanta, the most recent two-week case rate would nearly quadruple, according to the latest data analyzed by the AJC.
DPH also does not include demographic data such as race and ethnicity from antigen positive tests on its website, creating a gap in case information for a disease that disproportionately affects people of color.
Nancy Nydam, a spokeswoman for DPH, said the agency is working to include antigen data in its case rate calculations and county maps, but she did not provide a timeline.
Antigen tests considered ‘probable’
The bulk of coronavirus testing is done by reverse-transcription polymerase chain reaction, or PCR, tests and antigen tests. PCR tests, which detect viral RNA, are considered the most accurate.
But these tests are expensive and time and labor intensive to process. During the summer surge, PCR testing was pushed to its limits, leading to long waits for results.
Public health officials have put their hopes on cheaper antigen tests coming to market to help fill the gap, and tens of thousands of Georgians turned to antigen tests for results. Antigen tests detect proteins on the surface of the virus and can deliver results within minutes. They are most accurate when administered to someone with active symptoms of a coronavirus infection.
Fast results can help lead to faster isolation of infected persons and limit spread.
But there is a tradeoff for the speed. Antigen tests are less accurate than molecular PCR tests. They come with higher rates of false positive and false negative results, and for those reasons, antigen positives are considered “probable” cases, not confirmed.
The White House Coronavirus Task Force recommends states treat and report positives from the newest wave of antigen tests as new cases. In almost every way, DPH does.
When nursing homes, schools, jails and other providers report a positive antigen test, just like for the PCR test, the state’s policy is to trigger a contact tracing investigation and have patients quarantine, Nydam said.
“There is no question that antigen data provides critical surveillance about COVID-19 in Georgia,” Nydam said. “Like the PCR data, antigen data provides both outbreak information and informs mitigation efforts.”
Technical hurdles, DPH says
DPH computes the statewide rolling average and county case rate data with antigen cases on a weekly basis and makes it public — but only if you know where to look.
Only by digging deep into a separate County Indicator Reports website, linked at the bottom of the state dashboard, can one find weekly data on how new antigen cases are piling up in counties across the state.
The AJC analyzed the most recent County Indicator Report published Nov. 23. The newspaper’s analysis of the county-level data is likely an undercount of the severity of the discrepancy as the County Indictor Report provides data from Nov. 7 to Nov. 20, and a new one won’t be published until Monday.
Nydam said DPH has been hampered by several technical challenges in reporting antigen tests.
State and local health officials don’t always know which medical providers, nursing homes, schools and other institutions obtain rapid test machines in order to patch them into the state’s reporting system, she said. Many small providers that received antigen test kits also weren’t on the state’s electronic platform or hadn’t faced reporting such a large volume of cases before.
“Right now, data being reported may be incomplete or batched, causing daily antigen data to fluctuate widely,” Nydam said. “As DPH continues to work with providers and facilities, reporting will get better over time.”
Amber Schmidtke, a public health researcher and former Mercer University professor who tracks Georgia’s epidemic on her widely read blog, said the state has improved its data collection and visualizations over the course of the pandemic.
But the shortcomings with antigen test data are frustrating.
“So many people and organizations rely on (the state’s dashboard) to make decisions that affect millions of people’s lives,” she said. “It’s painting a very optimistic or rosy assessment of where we are.”
Schmidtke said DPH already collects and processes the data daily and could update the website to include antigen cases on the same charts and maps or create a toggle function to show PCR only and combined cases and case rates.
‘Information we need’
But another shortcoming is the lack of antigen test data in the state’s datasets of COVID-19 demographics. The dashboard includes cases and deaths by gender, age, race and ethnicity, but only for PCR tests.
The state has previously faced challenges collecting demographic data, which can help target health care resources by providing insights into which communities and age groups are suffering the most.
As antigen test use grows, the gaps in case data risk hampering the public health response to COVID-19 particularly in the most vulnerable communities, health experts say.
Schmidtke said cases, as determined by PCR tests, are skyrocketing for people 18 to 29.
“The demographic data is not just academic, it is actionable information we need,” she said.
Nydam said the agency plans to add antigen demographic data to the state’s dashboard. As more providers are added to the state’s electronic reporting system, “the antigen test data will improve allowing us to provide accurate and consistent antigen data on the dashboard.”
Staff writer and data specialist John Perry contributed to this report.
Find the latest information about the coronavirus pandemic in Georgia on the AJC’s coronavirus dashboard, including new data the AJC is charting on antigen positive tests.