Our coronavirus coverage is free for the first 24 hours. Find the latest information at Please consider subscribing or donating.

  1. News
  2. /
  3. Health

Ten questions about testing for the coronavirus

Head of BayCare labs explains challenges of COVID-19 process.

More than 3 million coronavirus tests have been processed in Florida so far. Some people have had more than one test, but even accounting for that, as many as one in seven people in the state have been tested.

Maura Pieretti, who oversees the laboratories for BayCare’s 11 hospitals across Tampa Bay, has 35 people in her labs who run tests for flu, MRSA, molecular diagnostics, genetics and chromosome analysis. Since the pandemic started, 10 employees have been testing for COVID full-time.

Pieretti’s labs have processed more than 39,000 tests in the last five months.

She answered questions from the Tampa Bay Times about the testing. The interview has been condensed and edited for clarity.

Maura Pieretti [ Courtesy of BayCare ]

What happens with drive-thru tests, from rolling down your window until the results come back?

The collector uses a nasopharyngeal swab, a very thin swab with a little fluffy head at the end, inserts it through the nostril into the back area of the nose and rotates it around a few seconds. Then the collector takes it out, puts it into a tube that contains a transport media, closes the tube, labels it and then sends it to the laboratory.

Tests on the respiratory sample basically look for pieces of genetic material that belong to the virus. That is done with a technology called polymerase chain reaction: PCR. It basically allows our instrument to recognize the genetic sequences that belong to coronavirus. Then we amplify them, so that we can easily detect them. There are a lot of chemicals and pieces in between, but it basically allows us to recognize the sequence and say yes, there is a signal, or not.

How long does that process take?

Our turnaround time depends very much on the location where the collection occurs. In some tests we do inside the hospitals, there's only a few hours in between. For testing we do in a centralized laboratory, it may be a day or two. And in cases where there is a backlog of too much demand and too little supply, it could take three days or longer.

How many different types of tests are there? Which ones work best?

If you go to the FDA website, there are several dozen different tests for the respiratory sample, looking at genetic sequences. … There has been a flurry of literature that has been published from March to today. And it keeps on growing, comparing one test to the other. I would say that all tests looking at genetic material are superior to tests that look, in general, at pieces of protein.

Antigen-based tests, which look at pieces of protein, are good early in the disease. But they’re not as capable of recognizing small amounts of virus. So they’re not as sensitive as the ones based on genetic material.

How accurate are the tests?

That's a very hard question to answer, actually, because accuracy is really made of sensitivity and specificity.

Sensitivity is the ability of a test to detect small amounts of virus. So tests are more accurate when they can pick up very small amounts of virus. The specificity of the tests that are based on recognizing genetic material is very, very high, because the tests are made to recognize only sequences that belong to the coronavirus.

There are very many components, though, in making a test. So if I do a collection on a patient that is in the very early stages of the disease and shedding a lot of virus, then my swab will easily pick up the virus. If I have a very small amount of shedding, maybe I'm in the late stages of the disease and the virus is in very small quantities, I may be picking up very, very small amounts of virus on my swab. And by the time I put it in the tube, it may be lower than the limit of detection. That's where the test may be inaccurate. It's just not sensitive enough for those very low amounts. Those are pretty rare and don't happen very often.

The other variable is where is the virus actually in the body when I'm trying to go collect it with a swab? The virus kind of hangs around the nose and throat at the beginning of the infection. Then it tends to travel downward, sort of in the back of the throat in that nasopharyngeal cavity, and then eventually is not even found there anymore. It goes into the lower respiratory tract. So depending on the timing when I do this collection, I may be able to collect virus or not. I may have a patient who has the disease, but I'm not capable of collecting virus from him anymore.

How often do you see false positives or false negatives?

False positive, in general, is a very, very rare event. It’s not something that is even being described in the literature very much, because the tests are very specific.

If somehow during the collection the sample gets contaminated, maybe by the collector or by somebody else, then it could be generating potentially a false positive result. Then there are the post-test variables. Was the result entered correctly from the instrument to the computer reporting the results? As we get busier, there’s the potential for more mistakes, like in any type of process.

False negatives actually happen more than false positives, from what is being reported. If I have such low amounts of virus that my test is just below the limit of detection, or maybe the virus already is not in the nose or nasopharyngeal cavity anymore, those could result in false negatives. There is not an exact number of how often that happens, because there are too many variables. The percentage of false negatives could be 20 to 30 percent.

Do you recommend people get a second test to be sure?

Not necessarily. We do retest patients if they are symptomatic and they need to be treated or they need to be isolated. If there is high suspicion that the initial test was a false negative, we can retest them.

Should people who don’t have any symptoms, but just want to know, go get tested?

My advice as a laboratorian is: Don’t get tested. We have such limited supply of tests still, and so few hands to do the tests. If you’re not symptomatic, don’t take those away from the people who really need them. Besides, to be tested is really only one point in time. You could become infected tomorrow. So the results are not very significant. It’s not very reassuring, either, in my mind.

What about blood tests that show whether someone has antibodies from having had the coronavirus? How accurate are those?

They’re very accurate in terms of picking up the antibody when it is present. So if it’s positive, it means the antibody is really there. If it’s negative, it means that it’s not there at the moment. But we don’t know if everybody is capable of forming antibodies after they’ve been infected. And we don’t know how long the antibody is present in patients. Or if it really protects against a new infection. There are some reports, unfortunately, of people not having antibodies anymore after a while.

Should people get tested for antibodies?

I think that’s not a bad idea because it allows you to know if you had the virus. And those are not just testing for one point in time. It’s more of your body’s reaction to the disease. Antibody tests are much more readily available than the other tests.

Do you have enough respiratory tests right now?

We could always use some more. It has been a struggle with manufacturers and vendors just being able to allocate supplies to us on a regular basis. They’re trying to meet demands from all over the country. So there’s a constant negotiation of how many tests do we need? Can they produce enough? Everybody’s ramping up production. But we’re also ramping up the number of tests we’re doing.

Times staffer Austin Fast contributed to this report