There’s been a lot of recent discussion and confusion regarding how Georgia reports their lab testing. I’ve tried to answer questions about it on Twitter, but I clearly need more than 140 characters to fully explain it.
There are two main elements that are at the center of the confusion – cases vs testing. We’re finding it’s pretty common for states not to require negative tests to be reported from all their labs, and some labs report only positive tests (ignoring the negatives). State law only requires labs to report positive tests, and this was never an issue with other disease reporting. However, with COVID, there has been a big emphasis on the percentage of positive tests.
In Georgia, we report confirmed cases from all sources, but only report testing data that comes from electronic lab reporting (ELR), where we have all the positives and negatives. For test reporting purposes, we just ignore all other testing sources. However, this makes our % positive appear higher than it actually is if you use cases divided by tests, because our testing numbers are admittedly incomplete.
This is the number of people with a positive PCR (viral) test for COVID. Cases are per person, not per test, and does not include people who tested positive for COVID antibodies.
Confirmed cases come from several sources: lab testing reported via electronic lab reporting (ELR), the State Electronic Notifiable Disease Surveillance System (SendSS), plus phone calls and faxes from various health care providers.
Only the ELR data includes both positive and negative tests (necessary to calculate % positive). The other cases may only be reports of those with positive labs, and thus would skew % positive. As a result, Georgia DPH chooses to only report ELR testing data. Cases that come from other sources do NOT have any testing data included in the testing section of the web site at all.
Note that as of the screenshot below on July 14, we have 123,963 confirmed cases but only 113,044 positive tests reported via ELR. That means at least 10,919 cases have come from non-ELR labs and thus their testing data is not reported at all.
This is the data that is reported via electronic lab reporting (ELR) only. No other sources for tests are included in these numbers. This means our number of total tests and positive tests are both under reported.
We can calculate % positive for these ELR tests by dividing the number of positive ELR tests by the number of total ELR tests (per day or overall). Unfortunately, we just have to assume that it’s representative of all testing being done (although we don’t have enough information to know if that’s likely true or not).
Tests are reported per test, not per person, so if all cases came from ELR testing, we should see fewer cases than positive tests every day (because some people will have multiple positive tests).
However, because we have cases that come from other testing sources outside ELR, confirmed cases for the day may be higher than the number of positive ELR tests. For example, in the screenshot above, you can see that on July 14, we had 3,156 positive PCR tests reported, but 3,394 new cases. This means that at least 238 cases came from non-ELR sources. However, it could be even higher (if there were duplicate positive ELR tests for people).
Antibody (Serology) Testing
Georgia also reports the total number of antibody tests have been reported via ELR, as well as the number of positive antibody tests, and the overall % positive for antibody testing. Just to reiterate, positive antibody tests are NOT included in confirmed cases.
Definitions from Georgia DPH
The following screenshots are taken from the GA DPH Dashboard to show my sources for the explanations above.