Infectious disease specialist Gonzalo Ballon-Landa, MD, of Scripps Health in San Diego, was a bit bewildered. One of his patients who had all the signs and symptoms of COVID-19 had tested negative with the nasopharyngeal swab RT-PCR assay. But when his hospital system's lab re-ran the sample, this time his patient was positive, just as he suspected.
And while Ballon-Landa is thankful his patient's infection was caught, he worries the test in wide use today may produce any number of false negatives. His guess: from 5% to 15%.
In a small who had fever or respiratory symptoms or had been in Wuhan or another endemic area, or had contact with people who had been there, only 71% tested positive by RT-PCR throat swab or sputum sample, while 98% had abnormal CT compatible with viral pneumonia.
In still another study that compared results by different tissue types , the nasal swab with RT-PCR testing revealed 63% of hospitalized patients with severe acute respiratory syndrome consistent with SARS-CoV-2 in China were positive.
That's critically important for a country that hopes to re-open the economy and get people back to work based on massive amounts of testing to assure a certified non-infectious healthcare workforce for all patients who need care, and to identify who should be quarantined or hospitalized, and who should be free to resume normal activities, said Colin West, MD, PhD, of the Mayo Clinic in Rochester, Minnesota.
40,000 Clinicians Infected but Negative?
In a report earlier this month , West and colleagues said false negative test results have critically dangerous implications for the safety of the healthcare workforce. Even assuming a 90% accuracy rate of the pharyngeal swabs most labs and clinicians are now relying on, "the magnitude of risk from false-negative test results will be substantial as testing becomes more widespread and the prevalence of COVID-19 infection rises," the report said.
In the U.S., West said, there are about 4 million doctors, nurses, and other clinicians providing direct patient care. If they are all tested at some point going forward, with an infection rate of 10% and a false negative prevalence of 10%, assuming that the testing gets much better than the earlier reports from China indicated, "there would be more than 40,000 false negative results," he said -- that's 40,000 infected practitioners treating patients.
"And if the sensitivity was not 90%, but was only 70%, as cited in some of these early reports such as those from China, the number of false negative results would triple, to well over 100,000."
That becomes important in testing large numbers of people, say in the state of California, with 40 million people, said study co-author Priya Sampathkumar, MD, an infectious disease doctor at the Mayo Clinic. That would mean 2 million false negatives if comprehensive testing were implemented. And if only 1% of California's population were tested, there could be 20,000 false negative tests just in that one state.
That means there would be a lot of healthcare workers treating patients while they were infected with COVID-19, and more waves of vulnerable people getting exposed and getting sick. That's also of concern as scientists raise new questions about whether getting infected with COVID-19 provides sufficient protection or immunity from getting either re-infected or having an illness recur from the original infection.
"We need to just be cautious about how we treat negative test results especially in higher risk individuals and we need better tests that I know our laboratories and our scientists are working night and day to get us," added West. "This is not a criticism of the test development industry or our universities; they're working really hard on this. But we need better tests, and we need to know how these tests perform so that we can interpret them properly."
Asked if he fears being perceived as an alarmist with suggestions that prevailing test strategies will send many infected doctors and nurses into hospitals, clinics, and practices to endanger patients, West said he's prepared for criticism.
"But I go back to the idea about which is the greater harm? Should we rather be overly cautious versus cavalier? The danger in a pandemic is much greater in being cavalier. So people who identify potential concerns that might lead to increased spread of infection should be taken seriously, and those concerns should be evaluated as we get more data."
Sam Torbati, MD, co-director of the Ruth and Harry Roman Emergency Department at Cedars-Sinai Medical Center in Los Angeles, had similar concerns, especially after reading papers suggesting the swab tests in use today have a 70% sensitivity. "That means that 30% of people who have the disease will test negative, and from a front-line standpoint, the concept that we have a test that will help us rule people in and out is not actually accurate."
What is the right time to test after a person is exposed but not sick? "We don't know at what point peoples' tests turn positive when they're not symptomatic. We don't have that yet in clinical science," Torbati said.
Bad Samples?
Many infectious disease experts said it's not known why some samples may be falsely negative, but it could be either the inadequacy of the sample collection -- the swab may not be collected properly or deeply enough to capture viral DNA -- or the tissue sample is not stored or transported properly.
"The obvious example, if a specimen is left in the back of a car for a day in a hot area and gets roasted, that's obviously not going to be good for a specimen," said David Louis, MD, pathologist-in-chief at Massachusetts General Hospital.
Angela Caliendo, MD, PhD, an infectious disease expert at Brown University in Providence, Rhode Island, added that "you have to go ... way back into the nasal pharynx in order to get a good quality specimen. Just swabbing the tip of your nose isn't going to give you as good of a specimen. We've also learned that just swabbing your throat isn't as good of a specimen."
Also important, Caliendo continued, is the timing of the test, and where people are in the course of their illness. "If you test me the first day I have symptoms, the amount of virus in my respiratory tract is going to be low compared to what it will be four or five days later. So it is the biology of the virus and the biology of the infection that also impacts how effective and how accurate these tests are."
Tom Frieden, MD, MPH, former CDC director, said recently that in New York, "we've seen a lot of false negatives" and there are "many bad tests, inaccurate tests on the market" today. "I think we're going to see kind of a Cambrian explosion and then a winnowing down of the tests out there so we have more tests that are reliable."
Frieden agreed that the difficulty of the nasopharyngeal swab plays a role in its accuracy issues: "Nasopharyngeal sampling is nasty," he said, because you have to reach far back into the nose and twirl it for 10 or 15 seconds. "That isn't always done, and if it isn't you can get a false negative."
Unknowns About Immunity
Another concern is the vast number of people who may carry the infection and be able to infect others who had no idea, as by scientists at Stanford University who surveyed 3,300 adults and children in California's Santa Clara County, where some of the first COVID-19 patients in the U.S. were confirmed. Using an antibody test from Premier Biotech that was not approved by the FDA at the time of their study, the researchers surmised that the prevalence of people with SARS-CoV-2 antibodies in that county, with a population of 1.9 million, could be from 48,000 to 81,000 in early April -- although the official case count at the time was just about 1,100. Nearly all of them would have no or very mild symptoms.
Equally concerning is that scientists aren't yet sure that after one has recovered from infection and presumably tests negative, whether that infection can rear up again. Deborah Birx, MD, White House coronavirus task force coordinator, acknowledged as much in an Sunday on Face the Nation.
"We just don't know if it's immunity for a month, immunity for six months, immunity for six years," she said.
Ballon-Landa said there is still so much that is unknown about how the COVID-19 virus behaves.
"We don't know this disease itself," he said. "There will always be outliers, in terms of people who shed the virus longer, and some who shed the virus a lot less."
The possibility of false negative tests is already causing worry in many healthcare settings, he said. "Now, one of the issues is, we're trying to get patients discharged to nursing homes and they won't take them. They are demanding two different tests be negative before they'll accept them."
No matter what, Ballon-Landa will wear a mask at all times when seeing patients, even if he's asymptomatic and tests negative. And time will tell whether he will gain greater confidence in whatever test for COVID-19 members of the healthcare workforce end up getting.
"Let's say you have a doctor who gets a negative test, but he had a positive test three weeks ago. And now we do a blood test that shows in fact he still has antibodies. Ask me then how I'll feel about it."