What do we know about school re-openings in Georgia?

A few weeks ago, I wrote a tweet thread summarizing the very informative “School Aged COVID-19 Surveillance Report” from Georgia DPH. Since then, updated reports have come out, along with some additional data downloads from GA DPH with age-based data, so I thought I’d review what we know now, after many schools went back face-to-face in Georgia the first week of August, and many more the first week of September (including Gwinnett County, the largest school district in the state).

Georgia Cases Epicurve

Cases peaked in Georgia in early-mid July, and declined steadily since, despite school re-openings. Cases are down over 60% across all age groups from August 1 – October 1.

Cases Among School-Aged Children

Statewide, cases among those <18 have been dropping since the July peak, with small bumps in early August and early September as many schools started back up. You can see that among K-12 students, the bump in cases around school openings was mostly among 14-17 age range. This is consistent with what we know about infections and transmission among children. Younger children seem to get and transmit COVID-19 less than older teens.

Cases among children are much less common than among 18-22 year olds. The graph below shows case rates per 100K people, so you can can see how cases among young adults far outpaced those among children throughout the pandemic. You can also see the size of the spike in college cases (18-22 years) compared to the minor bumps among 5-17 year olds as well as those 23+. Cases among the 18-22 age group peaked sharply as colleges started back in late August, but then declined sharply after.

Colleges & Universities

As you saw in the graphs above, cases among 18-22yo spiked in late August, as colleges returned, but then fell sharply to levels below that in early August.

This is consistent with DPH case data from counties with colleges. (These graphs are by symptom onset when available, otherwise lab collection date is used.) The counties that saw the largest spikes were Clarke (University of Georgia), Bulloch (Georgia Southern University), and Baldwin (Georgia College & State University). Cases now in these counties are already below where they were before schools re-opened in the fall.

This is also consistent with declines seen in reporting from the colleges themselves on surveillance testing as well as overall reported cases among students, staff, and faculty. The University of Georgia has stated that they have seen very few cases among faculty and staff. “The data suggest that spread is not happening through classroom settings, according to Dr. Garth Russo, executive director of the University Health Center and chair of UGA’s Medical Oversight Task Force” (source). In addition, wastewater analysis in Athens confirms saw virus detection peak in wastewater on September 1, and has already declined to levels “to or at the limit of detection.”

Percent Positivity

Percent positive has also declined among all age categories from 0-22 since August levels, so the case declines are not a matter of reduced testing among school-aged children. Percent positive peaked for 18-22yo in late July like it did for adults, then it hit a plateau in August as college students returned to campus, and then began declining again in mid-September. In 0-17 age groups, it peaked in early August (around school re-openings) then fell.

School Clusters

School “clusters” increased in August, and remained elevated in September. This is not surprising, nor particularly alarming, considering the broad definition of a “cluster”. It only takes 2 positive tests within 2 weeks among people who had contact at school. Nevertheless, clusters were lower in September than in August, and were down by half of the peak by late September.

It’s also a very small relative number considering even at the peak, it included ~40 clusters per week. Considering the large number of K-12 schools and colleges that have opened in Georgia, that is a very small percentage that have had even small clusters (of 2+ cases) since reopening. And note the very small number of day care center clusters. When discussing school cases and clusters, it’s important to keep in mind the denominator.

ER Visits

Kids also go to the ER – sometimes with COVID-Like (CLI) or influenza-like (ILI) symptoms. These graphs show the percentage of ER visits for CLI or ILI each age group. Visits for COVID-Like Illness have dropped significantly for children since the summer, while visits for Influenza-Like Illness have increased as we head into fall. Notice that ILI visits are highest for the youngest children, while CLI visits are higher in teens and college-age groups. This confirms what we know about flu versus COVID in young people, mainly that young children are at higher risk of infection and complications from influenza than COVID.

Hospitalizations Among Young People

Some children and young adults with COVID have been hospitalized – especially during the summer peak in Georgia. Fortunately, these are very low numbers overall, and a very low rate per case. Only about 1% of confirmed cases of COVID in children ages 5-17 result in hospitalization. Hospitalizations among teens and college age cases were at their peak during the summer months, before schools started back, and have declined since. Hospitalizations in younger age groups have remained at a very low level throughout. (Note that this graph appears to be by report date of the hospitalization, not the date of hospital admission, so the spikes and dips are probably an effect of reporting delays rather than dramatic week-to-week variations.)