Expanding access to rapid COVID-19 testing has been a federal priority for months, and over the fall, such tests have become much more accessible and popular in Louisiana.

The state Department of Health’s  recently began including rapid antigen tests — along with confirmatory PCR tests — in its COVID-19 data dashboard. The antigen tests are included in the state’s daily case counts, albeit with a disclaimer. Positive antigen tests are considered “probable,” not confirmed cases. On Tuesday, the state reported 2,429 new cases, of which about one-quarter were considered probable cases from antigen tests. On Wednesday, more than one-third of 4,339 newly reported cases came from antigen tests. 

Antigen tests aren’t calculated into positivity rates, one of the main metrics the state and local governments have used to make decisions on mitigation measures, such as restricting business capacity limits. Considering the growing number of rapid tests being given and the fact that they aren’t considered in the official positivity rates, what do those rates mean now?

The answer isn’t a disaster, according to local public health experts. Rapid testing opens up new opportunities for the state, but to do so, it needs to figure out where tests are being conducted.

PCR tests are considered the gold standard for COVID diagnosis, with relatively low rates of false negatives and positives, and are what the city of New Orleans offers at all of its community testing sites. However, the tests take days to return results because they have to be processed in a large central lab.

Right now, there are at least four rapid COVID tests available on the market. The two most common, BinaxNOW and ID NOW, are both manufactured by Abbott Laboratories. After the FDA approved BinaxNOW for emergency use, the federal government purchased 150 million of the tests, and began distributing them directly to states and healthcare facilities like nursing homes. It has also purchased and distributed ID NOW tests at a smaller scale.

Rapid tests function slightly differently from PCR tests, and are all described as “rapid” only because they return results quickly without the need to send samples to an outside lab. BinaxNOW is a bit like a credit card-shaped pregnancy test and tests for fragments of viral protein, called antigens, in saliva. ID NOW is a small machine that tests for the presence of viral genetic material, like a sped-up PCR test, and is referred to as a “molecular test.”

Louisiana has received close to one million BinaxNOW tests from the federal government, with 575,000 sent to LDH, and another 400,000 delivered directly to nursing homes and other facilities.

LDH hasn’t actually distributed all of those tests, although it didn’t answer questions about how many it has left.

But that one million tests likely doesn’t begin to approach the total number of rapid tests in the state. Walgreens and CVS both offer rapid testing in Louisiana, and individual clinics may have their own purchasing agreements separate from the federal distribution. Loyola University also used rapid tests for diagnosing students.

From the outset, rapid tests, and especially antigen tests like ID NOW, have faced questions about their false negative rate, or sensitivity.

Because of that, according to an LDH guidance to healthcare providers, antigen tests like BinaxNOW are not recommended for use in asymptomatic people with no suspected COVID exposure.

“Rapid antigen tests perform best when the person is tested in the early stages of infection with SARS-CoV-2 when viral load is generally highest,” the guidance reads. The guidance leaves some room for discretion on the part of healthcare providers, suggesting that “if a person with no symptoms and no known contact with someone with COVID-19 tests negative with an antigen test, they should be counseled on the limitations of antigen testing in this setting.”

That’s not to say that the less-sensitive rapid tests are worthless. One school of thought, articulated by influential Harvard epidemiologist Michael Mina, argues that the benefits of catching and isolating cases quickly outweighs the downside of false negatives.

That’s why LDH leaves the usage of rapid tests up to provider discretion in the end, and why some nursing homes in Louisiana and across the country have used rapid tests for surveillance of all staff and residents.

“Given that that environment is such a high threat,” said Susan Hassig, a professor of epidemiology at Tulane University and a member of the state’s COVID-19 task force, “being able to test almost constantly is a strategy that will work both for the staff as well as for the residents. The challenge is in places outside of congregate settings and clinical settings. The bottom line is, whatever your test result is you probably should quarantine.”

But does anyone know how many tests are being reported?

There’s a clear distinction between the reporting process for PCR tests and antigen tests. Because PCR tests are processed at large central labs, like LCMC or Tulane Medical Center, there’s a longstanding process for reporting test results directly to LDH.

“They have systems that are automatic for everything,” Hassig said. “Syphilis, gonorrhea, measles, everything that they do diagnostics for.”

Rapid tests have to be performed by an institution with a baseline level of FDA certification, but they don’t necessarily go through labs with an established reporting process. Instead, individual clinics, pharmacies, school districts, universities, and other providers are responsible for reporting tests to LDH.

Technically, they’re supposed to report all rapid tests, positive or negative, to LDH within 24 hours. For many of the institutions, it’s the first time they’ve ever had to report to LDH, and to do so, they need to log into an LDH portal and enter results manually.

LDH has leverage over the institutions, like school districts and HBCUs, that it provides rapid tests to, because it can simply stop sending more tests, and so it has some sense of rapid testing numbers. But in practice, it doesn’t have a way of enforcing reporting from providers that get tests from other sources, and so it doesn’t know exactly how many rapid tests are being conducted in Louisiana.

“The federal government pushed out these kits and machines without telling LDH,” Hassig said. “They provided rapid tests to nursing homes, but they neglected to have LDH in on the loop.”

For the last month, LDH has displayed the total number of known antigen tests conducted on its COVID dashboard. As of Tuesday, that total was 125,000 antigen tests, compared to 3.6 million molecular tests. (ID NOW tests are reported in the antigen category, Hassig said, despite functioning slightly differently.)

According to seven day averages of that data, antigen tests have made up about 10 percent of the state’s reported test volume for the past month.

But because of its uncertainty over the extent of rapid testing, LDH does not incorporate rapid tests into its calculation of percent positivity.

Of particular concern, Hassig said, is that negative tests may not be reported consistently, and so the total number of tests would be “unstable in a whole new way.”

“They may think, oh I need to tell them when I find somebody positive, but they may not be attuned to the fact that they need to tell them how many negative tests they’re doing too.”

The reporting gap between the two types of test may become more important as time goes on, because demand for and availability of rapid tests appears to be rising. According to a report from Emily Woodruff in the Times Picayune | Advocate, visits to LCMC’s New Orleans urgent care clinics, which provide rapid testing, increased 30 to 50 percent in the two weeks before Thanksgiving.

At the same time, public schools like Bricolage Academy are starting in-house rapid testing programs. Ochsner urgent care clinics currently provide walk-up molecular rapid tests to asymptomatic people, but in a statement sent to the Lens, a Ochsner spokesperson wrote that the tests were available “on a limited basis … for travel and other circumstances where a test is required.” The hospital system did not answer questions about the extent of that testing, its reporting practices, or how it advises people to interpret test results.

LCMC did not respond to questions about its rapid testing practices, but an FAQ page on its website notes that a patient must be “screened by one our urgent care providers” to receive a rapid test, and that some people may not be eligible for rapid testing.

It’s not obvious how incorporating the rapid testing data might change percent positivity numbers. According to both Hassig and Susanne Straif-Bourgeois, a professor at LSU who studies infectious disease epidemiology and also serves on the COVID task force, it’s likely that the rapid tests would actually decrease the state’s overall percent positivity. Many rapid tests are being used to screen asymptomatic staff at nursing homes, where the higher false negative rate would lead to a lower percent positivity.

But in other settings, like the New Orleans Public School District, rapid tests are being used to diagnose people who are already showing symptoms, per the LDH guidance. If the tests are being used on large numbers of people who have symptoms, they’re likely to have a higher overall positivity rate. And if sick people are diagnosed by rapid test and never receive a PCR, those positive tests will never enter the official percent positivity numbers.

Percent positivity

Percent positivity isn’t the only data point public health officials use to assess outbreaks, Straif-Bourgeois pointed out.

“I still believe that percent test positivity is a good metric, but you will have to look at all these metrics combined, such as the number of cases over time, hospitalized COVID-19 patients, and patients in ICUs.”

Percent positivity has been leaned on heavily by parishes and the governor in making decisions about COVID precautions, including business shutdowns and reopenings. Gov. John Bel Edwards’ Phase III decision, for instance, allowed bars to reopen in parishes that stayed below five percent positivity for two weeks.

But Hassig said that percent positivity is only a rough tool for measuring the extent of an outbreak.

“A high test positivity means that you’re not testing enough to know how many people are infected in your population. If you get above 5 percent it’s getting kind of tricky to feel like the cases that you’re identifying are really giving you a good sense of how big a problem is,” she said. “If you’re above 10 percent, you’re not testing nearly enough. You’ve got a lot more people out there that are infected that you’re not finding.”

Because of that, Hassig said, she isn’t concerned that rapid testing is being left out of the number, although she thinks it would be appropriate for LDH to publish positivity for both rapid and PCR tests separately.

This isn’t the first time that LDH has had to develop new reporting relationships during the pandemic, though, she said. “Over the summer surge, there were new locations that had not previously been reporting PCR data to the state, and there were all kinds of issues getting that information integrated. It’s a major undertaking.”

A larger issue with the lack of data about rapid testing is that it represents a missed opportunity for the state to control its outbreaks, Hassig said. 

“We have to test to find [infected] people, but just counting the people positive is not an end in and of itself.”

The goal, she said, should be getting people into quarantine or isolation and breaking chains of transmission. This past week, 56 percent of positive COVID cases in Orleans Parish were not contacted within 24 hours.

“Theoretically,” she continued, “you could get the result in 15 minutes [with a rapid test], you could tell the person right there that you need to go home and isolate, and that you need to tell us right now–before you leave the office–who you’ve been in contact with in the last 48 hours. It could all be one package.”

What that would take, however, is clear channels of communication between LDH and the clinic performing the test.