State COVID data ’opaque’ as infections rise in Georgia. What do the numbers really mean?

On June 22, the Georgia Department of Public Health reported 538 new COVID-19 cases in Georgia with a seven-day rolling average of 908.

On the same day, the Atlanta Journal-Constitution reported 1,073 new cases and a seven-day rolling average of 1,227, while the respected Johns Hopkins University COVID-19 tracker showed 1,882 new cases and a three-day rolling average of 1,334.

Which numbers can you trust?

Richard Griffiths of the Georgia First Amendment Foundation asked panelists during a recent online forum the foundation organized, “Deciphering Public Health Data about COVID-19.”

“It comes down to dates, and how you attribute dates,” said panelist Ben Lopman, an Emory University epidemiologist.

A case could be entered by the date of onset; the date a person sought care; the date someone got tested; or the date it was reported.

Other panelists were Atlanta Journal-Constitution investigative reporter Willoughby Mariano and Andy Miller, CEO and Editor of Georgia Health News. The Georgia Department of Public Health was invited to send a representative but declined.

The Georgia Department of Health (DPH) COVID data tries to get at that earliest date, to reconstruct how the COVID-19 pandemic has unfolded, while the other two are more in the present, Lopman said.

“If you want to get a better sense of what’s going on at any given time, what the trends are, the simplest metric is what’s being reported,” he said.

One thing the recent data shows is that the number of infections is climbing, in Georgia as in many other states, and it’s not just because of more testing, Lopman said.

“It’s quite clear that the increase you’re seeing is not a result of increased testing. The percent positives is increasing,” Lopman said as Griffiths displayed a Johns Hopkins graphic showing the percentage of June 22 test reports that were positive on June 22 in Florida (13.0 percent, trending up) and Georgia (8.9 percent and trending up.) Georgia is just below the 10 percent threshold the state of New York set for requiring visitors from those states to quarantine for 14 days, with fines up to $10,000 for violators.

Hospitalization numbers are also a good indicator, but the best single indicator is deaths, Lopman said.

There’s less bias—you know when someone has died—but death reporting has its own tangles. Even without COVID, assigning a cause isn’t always simple when multiple health factors are in play.

Another state agency, the Georgia Emergency Management Agency (GEMA) tracks an overlapping set of COVID statistics, including hospitalizations and the availability of critical care beds in each of the state’s health based Northeast Health District.

On Thursday, GEMA’s daily Georgia Situation Report showed just eight critical care beds available in the Athens-centered Northeast Health District, with 62 in use. That was the lowest number available since the state began making the statistic available to the public April 17, according to retired University of Georgia journalism professor Lee Becker. Becker has also been tracking the pandemic and paying particular attention to the Northeast District, which includes Clarke, Oconee and eight other nearby counties.

As of Thursday, Georgia COVID hospitalizations had gone up 12 straight days, he noted in his Oconee Observations blog. Hospitalizations rose again Friday, when the Department of Public Health reported 1,900 new confirmed cases—the highest one-day total yet, Becker found.

Statewide, critical care bed availability had declined to 26 percent Friday.

“This is one of the real concerns about this pandemic,” Lopman said in the First Amendment Foundation discussion, “to be able to respond, to have the capacity to treat patients.”

One place where cases are on the rise is in Clarke County, which had recorded 463 confirmed cases since the beginning of the pandemic—up 89 in just the last week.

The COVID death rate in Clarke County has been low so far. More than 40 other Georgia counties have recorded more deaths than Clarke County’s 15 as of Friday, most of them counties smaller than Clarke. Oconee County is now up to 10.

Death rates have not been rising to match infection rates, but that might change.

Death is the most objective indicator, but it’s the worst indicator of what’s happening right now, Lopman said. It’s a lagging indicator, with an average gap of about two weeks from hospitalization to death.

A total of 2,770 people have died of COVID in Georgia, according to the DPH, and more than 124,000 nationwide, according to the CDC.

But those are certainly undercounts.

“Early on there was not much testing,” Miller said. Persons who died of pneumonia in February could have gone unrecognized.

Some deaths not counted as COVID are likely an indirect effect of the disease, such as people avoiding seeking medical care for chronic conditions during the pandemic.

Now there’s evidence that more young people are testing positive, which is likely to dampen death rates since younger people are less severely affected by the coronavirus than older people.

Usually during a crisis, the media has access to crisis response centers so the public can know how officials are making decisions.

That’s not been the case during the pandemic, Mariano said.

“They’ve not let us in,” she said.

“This has been an opaque process, and I can’t think of a time when we need more information to make our own personal decisions,” Miller said. “We ask questions, and a lot of times we just don’t get any answers.”

Another problem is that our data collection systems just are not very good, Lopman said.

“We just have outdated systems considering this is the 21st century,” he said.