The Week in Review, 02-08Nov
The Week in Review, 02-08Nov
Georgia COVID-19 Updates
This week we started getting daily updates on cases identified through antigen testing. They’re making up a significant proportion of each day’s cases, ranging from 18 - 38% of newly reported cases each day. However, since DPH is only graphing cases, testing, hospitalizations and deaths based on PCR test-identified cases, we are not getting thorough demographic data. If the DPH is treating each antigen-test identified cases as cases, each initiating a contact trace investigation, then we really need to get age, sex, and race/ethnicity data for those cases too. Otherwise, it’s hard to know where our increases are happening and how to best target testing and messaging efforts to get ahead of the pandemic. Hopefully we start seeing those data at some point soon. The other problem with the state only graphing PCR-based cases is that it means their case curves show increases but don’t look so bad. If you include the antigen-based cases, then our case rate is increasing sharply.
Combined Data
Let’s begin with how things are trending for three of our key indicators. The blue line represents cases and corresponds to the left y-axis. The gray line represents hospitalizations and the red line represents deaths and these lines correspond to the right y-axis. I graph these together because it can be helpful to see how the three are linked to each other over time. For example, with the summer surge cases spiked first followed by hospitalizations. But there was a 3 week delay before deaths began to track upward.
This week we see increases across cases, deaths and hospitalizations. This graph shows cases identified through both PCR and antigen testing, starting with the most recent week - as in, the total cases today (PCR + Antigen) minus last Sunday’s total cases (PCR + Antigen). The problem is that last Sunday, we were only being given the antigen cases weekly. So I calculated cases based on what they told us the previous Monday for antigen cases and then adding each day’s newly reported PCR cases to that total. So last Sunday’s “total” case count was likely an undercount because I only had Monday’s antigen test data. I think that’s why this week’s case line shows such a large increase. I think we should be cautious but not panic. Some of this large increase is an artifact of imperfect data. Next week will be more reliable. This week, antigen testing comprised 41% of the total cases. So their impact is significant and you see the sharp increase in the blue line this week. We’ll have to see if that trend continues. But antigen testing are already making a big impact. A reminder on how these tests work - they aren’t very good at picking up low amounts of virus. The PCR test is the best for picking up small viral quantities of virus. So we need to confirm antigen negative tests with PCR. So the positive results for the antigen test are more trustworthy than the negatives. I’ve confirmed with DPH that each antigen case initiates its own contact trace investigation.
We see smaller increases in hospitalizations and deaths. These small increases have been observed before, not really materializing into any sort of meaningful surge. So we’ll see what, if anything, they mean next week and in the weeks to follow.
Testing
This week’s testing output higher than the previous week. Meanwhile, the percent positive rate decreased slightly. Keep in mind, these data are based only on PCR tests. It does not factor in antigen testing.
In total, there were 149,849 new test results reported this week through Electronic Laboratory Reporting. Of those, 6.9% were positive.
Cases
The number of new cases (by date of report) is as follows:
- 24 hours: 1439 (+0.4%)
- 7 days: 18,814 (+12.2%)
- 4 weeks: 74,652 (+22.5%)
These are larger numbers than we have seen in recent weeks, mostly because of the inclusion of antigen test-identified cases. The new statewide total is 406,061, combining PCR and antigen-test identified cases. Bear in mind that these case totals are likely an under count of our actual case count - it’s a problem that plagues public health for multiple diseases and especially so when a disease has so many asymptomatic cases, as COVID-19 does.
The graph below shows how each week’s cases have been distributed by age according to PCR-based testing alone. Each bar represents a separate week. This kind of graph lets us look at timelines for each age group at the same time. Note, the Department of Public Health does not yet include antigen-identified cases in their demographics data table or graph. This week, antigen-identified cases made up 18-38% of each day’s newly reported cases. So there is a lot of data we aren’t seeing. But the graph below shows what we *do* know, based on PCR-based testing.
This week, cases remain high among those 18-59. Increases were observed for those 10-17 and 60+. If we look at a similar graph below that shows the timeline for different races, we can see that increases were noted for the Asian and American Indian/Alaska Native category. Again, this graph is based on PCR-identified cases only. Antigen test-identified cases are not included in these race totals provided by DPH.
The case rates per 100,000 people are increasing across all county types. The statewide average (black line) is trending up sharply. Ignore the weird peak around 06Oct. That’s when DPH first started providing antigen test-identified cases, dumping all of the cumulative cases to date. Pay attention to the trends that happened after. The statewide case rate is 23% less than where we were at the height of the summer surge on 28Jul. As discussed in the introduction, this graph is going to look different than the graph provided by DPH because I’m factoring in the antigen test-identified cases.
The other thing to note is that the increase we see here is happening across all county types. There isn’t really one county type that stands out compared to the others. I’ve mapped 14-day case rate per 100,000 for each Georgia county below, using both PCR and Antigen test-identified cases. I recommend clicking on the map to see the live image where you can click or hover over a particular county of interest to get their value. The color for 200 cases per 100,000 (a threshold set by some school districts for when to pivot to virtual learning) is right at the transition from a salmon color to magenta. You’ll notice that many Georgia counties are at or above that threshold.
Hospitalizations
There were 748 Georgians newly admitted to the hospital for COVID-19 this week (+15% compared to last week) and 134 were admitted to the ICU. The number of patients currently hospitalized (shown below) has been trending up but hasn’t risen sharply yet like we saw during the summer surge. Hopefully that history doesn’t repeat here.
If we look at how hospitalizations have varied over time with age, we can see that this most recent week featured both increases and decreases in hospitalizations for many age groups. Hospital admissions increased for age groups 5-9 and all those >40 years old. Again, this only considers hospitalizations among cases identified through PCR.
Here’s how new hospital admissions have varied for different races over the past month. We see increases this week for Black / African-American, White and Asian populations.
ICU admissions (below) show decreases for many races (a good thing), but we do note an increase for those who identify as White or Asian.
For our healthcare workers, there were 648 newly reported cases, 25 hospital admissions and 2 newly reported deaths this week. A reminder, our healthcare workforce is not an unlimited or renewable resource. Please limit your exposures to honor the risk that our healthcare workers are taking to care for the ill right now.
Deaths
The number of new deaths (by date of report) is as follows. The percentage in parentheses indicates how the increase compares to Georgia’s cumulative total at the start of each window of time.
- 24 hours: 1 (+0.01%)
- 7 days: 213 (+2.7%)
- 4 weeks: 778 (+10.5%)
This week the number of deaths was an increase of 41 compared to the week before (172 deaths). The current statewide total is 8194 and of those, 9.5% were reported in the last 4 weeks alone. The case fatality rate is 2.20% if PCR-based cases alone are considered. If the antigen test cases are also considered, then the case fatality rate is 2.13%. The case fatality rate has been falling for the past three weeks with the increase in cases to dilute the deaths.
As for cases and hospitalizations, here is a graph of deaths by age over time. As you get to the next age group, the timeline starts over. A reminder, this graph is built using data for deaths among PCR-identified cases only.
This week we saw increases in deaths for every age group over 40 years of age. Deaths are a lagging indicator compared to cases, usually by about 2-3 weeks. So as we see cases increase across the state, we might expect the deaths increase this coming week or next. Hopefully they don’t. But that’s what’s happened in the past. Next, let’s look at how the most recent week’s deaths were distributed by race.
Increases were recorded this week for those who identify as Black / African American, White, Asian, and American Indian / Alaska Native. Below you will find the updated 7-day death rate per 100,000 graph, showing how deaths have varied across county types. As noted elsewhere, death rates increased ~15% compared to the previous week. The rural death rate remains 82% higher than the state average. The death rate has ticked up across all county types but less so for the Atlanta counties.
References
https://dph.georgia.gov/covid-19-daily-status-report
https://covid-gagio.hub.arcgis.com/
Georgia COVID-19 Updates is a free newsletter that depends on reader support. If you wish to subscribe please click the link below. There are free and paid options available.
My Ph.D. is in Medical Microbiology and Immunology. I've worked at places like Creighton University, the Centers for Disease Control & Prevention and Mercer University School of Medicine. All thoughts are my professional opinion and should not be considered medical advice.