Skaman306—Getty Images
May 8, 2020 8:00 AM EDT

In late February, several weeks before the coronavirus outbreak shut down American cities and rose to the level of a national crisis, Kerri Rawson began to feel sick. “I was hit out of nowhere with what feels like the flu at first,” says Rawson, who also has asthma and takes cardiac medication for high-blood pressure. “You’re fine, and then all of a sudden you have a fever below 100°F and chest congestion.”

Rawson is a 41-year-old writer and mother of two in Florida. (You may recognize her name from her 2019 memoir, about growing up as the daughter of a serial killer.) Her fever lasted for 11 days, during which time her children also developed above-normal temperatures. Her son’s fever rose to 102°F but tapered off in a few days; her daughter, however, developed a barking cough that Rawson had never heard before. A doctor diagnosed the 11-year-old with bronchitis.

“That’s when my first conversation about COVID-19 started,” Rawson says. On March 6, still struggling with fever and chest congestion, she asked her family doctor whether she might have contracted the coronavirus. He was skeptical. (There were, at that point, fewer than 10 reported cases in Florida.) “He asked me questions about traveling and contact,” Rawson says. “He said, ‘Our hands are basically tied by the CDC [the U.S. Centers for Disease Control and Prevention]. We can’t test. Call the state health department, call the local one.’” Rawson did so, but was told they were only administering tests to those who had traveled internationally, had contact with someone who had, or were in critical condition, none of which described her.

Over the following week, Rawson’s condition worsened. Her blood pressure rose, her heart rate was up, and she began to have shortness of breath. Rawson saw a family doctor. “I just sort of collapsed on her table,” she says, and told the doctor she was worried it might be COVID-19.

The doctor sent her to the emergency room. “I was basically in hypertensive crisis by the time I got to the ER,” Rawson says. She was admitted to a hospital in Altamonte Springs, FL, and placed in a room on an observation floor with a sign on the door requiring nurses to take precautions like mandatory gloves and surgical masks. Doctors tested her for “everything under the sun,” she says, but not COVID-19.

Courtesy of Kerri Rawson
Kerri Rawson

It had now been nearly two weeks since Rawson first noticed any symptoms, and she still had not been tested for the virus—a sadly common tale during the early weeks of the pandemic, when U.S. officials overwhelmingly failed to make widespread testing available to sick Americans. Florida’s pandemic response, in particular, was compromised by meager funding to state and county health agencies and cuts to research funding, according to a Tampa Bay Times investigation, as well as Governor Ron DeSantis’s slowness to issue a stay-at-home order.

On March 12, Rawson received a CT scan. When doctors saw the results, “they freaked out,” she says. She was diagnosed with bilateral pneumonia. Most concerningly, the scan of Rawson’s lungs revealed “ground-glass” opacities—abnormalities in the lungs that show up as grayish patches, resembling ground glass—that are common among COVID-19 patients. “When they saw the ground-glass look in the lobes, they contacted infectious disease, and that’s when everything hit the fan,” Rawson says. “Friday morning, the nurse comes barreling in, tosses all my stuff on my bed. They throw a sheet over me. They put me in the hallway, they wipe down my bed, put a mask on me, and rush me through a couple floors up to the zero-air containment room.”

Finally, on March 13, after being moved to an isolation room on a progressive care floor and prescribed two different antibiotics, Rawson received the nasal and throat swab test for COVID-19. Six hours later, the test came back negative.

Rawson believes it was a false negative, and that the test was not administered correctly. “I ended up having a really bad nosebleed and my swabs were covered in blood,” Rawson says. “[A nurse] in the ER said that could have even affected the test.” (We’ve reached out to the hospital, AdventHealth Altamonte Springs, for comment in response to Rawson’s claims in this article. The hospital has not provided an on-the-record comment.)

Soon she was kicked out of the isolation room and moved back to the observation floor. “They ended up having to evict me at like 1:00 a.m. because they needed it for someone else,” she says. “And the night nurse didn’t really want to be around the COVID [patients]. She wasn’t really having any of it. I had to, like, push my dumb IV pole around and collect all my stuff when I was really sick.” Rawson was told the room was needed for another suspected COVID-19 patient.

On March 14, she went home, where she spent a week battling a difficult recovery from pneumonia, including suffering from neurological issues and sleep deprivation—“it was horrible,” Rawson says. She wound up back in the ER a week later when her fever returned. By mid-April, she still had not fully recovered.

Reason to be skeptical of test results

Rawson’s experience with the virus—assuming this was indeed COVID-19—was extreme, but her testing experience is not uncommon. The nasal swab diagnostic test, which involves amplifying small traces of DNA using a laboratory technique known as polymerase chain reaction, or PCR, is far from infallible. One preprint article from China estimates the false-negative rate to be as high as 30%.

In practice, that figure would mean that “if you tested 100 people who all had COVID-19, 30 of them would still get a negative result,” says Dr. Catherine Carver, a PhD student in Population Health Sciences at the Usher Institute, University of Edinburgh.

This would also mean that thousands of Americans have received test results telling them that they do not have the virus when in fact they do. “This is a significant problem because it could create false reassurance for the people getting the false negative result that they are well and won’t infect other people,” says Carver.

In early April, a Yale physician grew alarmed and wrote a New York Times op-ed urging patients who have coronavirus symptoms but test negative to assume they are positive. Citing anecdotal evidence from fellow doctors, he noted that such situations are “uncomfortably common.”

So far, there is little reliable research into overall COVID-19 test performance. But it’s dangerous to place too much faith in the test’s verdict, says Dr. Colin West, a physician and professor of medicine at the Mayo Clinic in Rochester, MN. “Testing is still going to be a very important part of managing this pandemic,” West says. “But we need to understand that the tests aren’t perfect. No test is perfect. And if there’s a certain percentage of false negative results that we may expect, we need to be cautious and not celebrate too soon if we get a test result that comes back negative.”

West is the co-author of a recent article in Mayo Clinic Proceedings warning against over-reliance on COVID-19 testing. Even if the test is 90% accurate, the paper states, “the magnitude of risk from false test results will be substantial as the number of people tested grows.” (Suppose 5 million people are tested; that could mean 500,000 false results.)

That doesn’t mean testing isn’t a crucial element of the nation’s pandemic response—or that the administration’s failure on this front is anything less of an outrage. On a population level, mass-distributed tests will be essential to mitigating the crisis. It does, however, mean that patients and care providers alike need to resist the temptation to regard test results as gospel.

There are several reasons the test may deliver inaccurate results. The first is that a nasopharyngeal swab is simply not easy to perform. “Doing it properly requires sending the swab fairly deep back into the nasopharynx,” West says. “There has been concern that, in some cases, the swabs are not getting back as deep as they need to go. The nose is being swabbed instead of getting to the back of the throat.”

Another possibility is that—depending on timing—a patient who has the virus may not have it in sufficient quantities for the sample to render a positive test result. “It turns out that the viral load and the performance of the PCR tests actually drops after a number of days of symptoms,” says West. “So if you wait too long, you might get a false negative. But if you do it too early, you might get a false negative as well because there isn’t enough viral material.”

While ramping up public testing has been a priority, experts say there also needs to be more research into test reliability. “Doctors and patients need to know how much faith to have in these tests, so they can make the right decisions about patient care or safely going back to work,” says Dr. Carver, who is the co-author of a recent paper arguing that there is not yet enough data to accurately assess COVID-19 test accuracy.

Emotionally, a negative test result can also cause more stress and uncertainty for suspected coronavirus patients. “It has been absolute hell,” says Eva, a Los Angeles-based music producer who tested negative twice after getting sick in mid-March, and who prefers to keep her last name confidential for privacy reasons. “I’ve had two doctors tell me I probably don’t have COVID, and some say I probably did have it. I’ve been going crazy calling doctor friends and asking for help, thinking I have cancer or a blood clot or something.”

Turned away from the doctor

Meanwhile, Kerri Rawson still doesn’t know for sure if she had COVID-19.

What she does know is that the possible false negative has made it more difficult for her to receive the medical care she needed.

On March 22, Rawson returned to the emergency room with cardiac issues. By then, the outbreak had risen to the level of a national emergency. The hospital now had a more intense protocol for patients displaying symptoms, Rawson says: “If you flagged for possible COVID, they didn’t put you through triage. They basically sent you back to a zero-air room and triaged you there.” (This type of isolation room “had its own air system that wasn’t attached to the greater hospital,” Rawson explains.)

Rawson was tested for the virus a second time. Around then, Rawson says, a nurse told her that the hospital staff had been retrained on how to administer the swabs since her last test. “They told me they had been having a ton of false negatives like two weeks before, when I was in there,” Rawson says. “And now they’re supposed to swab up and hold for three seconds, which they weren’t doing before.”

According to Rawson’s account, the nurse said patients were previously being tested on one floor, receiving a negative result, then being tested on a different floor with a positive result. Rawson also heard that the hospital had instituted a new policy: If someone was tested due to COVID-19 symptoms and the test came back negative, they would be tested again 24 hours later.

Rawson was told to expect the results of her second test in two weeks. It’s been more than a month. She never received them.

That became a problem when her family doctor advised her to follow up with a pulmonologist, given her asthma condition. She tried to make an appointment. “I mentioned I had been in the hospital with what doctors were saying was highly clinically significant for COVID,” Rawson says. The office demanded proof that Rawson had either never had COVID-19, or no longer had it. She suggested a virtual appointment, but the office had a policy against arranging telemedicine appointments with new patients. “I basically have to find a new pulmonologist if I need one,” she says. Unless she could provide a second negative test result, “they were outright refusing to see me.”

Rawson plans to see if her doctor can order her an antibody test, which is meant to reveal whether the patient has built immunity to the virus. (Such tests may also not be entirely reliable.) She has felt vindicated by reporting from the Palm Beach Post suggesting that COVID-19 may have infected hundreds of Florida residents as early as January or February, long before the state acknowledged its first presumed cases. Meanwhile, the state’s attempts to control the outbreak have been stymied by a massive testing backlog, which could include Rawson’s second test.

By late April, Rawson had mostly recovered, but still felt some fatigue. Her lungs remained weak. She could barely make it up a flight of stairs without stopping to catch her breath. She has no idea if she will ever receive the results of her second swab.

“I’m assuming they sent it off,” she says. “I have no record that they sent it off.”

Contact us at editors@time.com.

Read More From TIME

Related Stories

EDIT POST