As COVID-19 infections continue to spread, demand for diagnostic tests has risen to a level that has pushed back California's mass-testing expansion plan due to a strain on resources. Now, "specimen pooling" could be a game changer, making hundreds of thousands of tests accessible to the public.
Dr. Sonia Angell, the state's public health officer, rolled back testing guidelines on July 14 to essential workers, those with symptoms of the virus and people who were in close contact with an infected person or who were in high-risk settings such as public transportation and mass gatherings due to a shortage of testing components and testing-result delays at laboratories.
The rollback puts a damper on the widescale testing and retesting that health experts say is crucial to control the outbreak and to protect the economy. Nobel Prize-winning economist Paul Romer, formerly chief economist of the World Bank, said in his proposal "Roadmap to Responsibly Reopen America" that to save the economy and lives, everyone in the United States should be tested every two weeks. Patients with positive results should self-isolate while the economy reopens.
If that's the case, then the U.S. has a long way to go. Under Romer's roadmap 7% percent of the population would need to be tested every day, David Donoho, Stanford University professor of statistics, said in a recent SIAM News article. Donoho and his Stanford colleagues, Mahsa Lotfi and Batu Ozturkler, have researched worldwide emergent mathematical and statistical ideas regarding COVID-19 specimen pooling published July 13 in the publication by the Society for Industrial and Applied Mathematics.
"For context, only about four percent of U.S. residents were tested for COVID-19 from March through May," he noted.
Specimen pooling, also known as group testing, can vastly increase the number of tests, save scarce resources, reduce costs and speed up results, he and his co-authors said. The method received a boost last weekend after the U.S. Food and Drug Administration authorized one laboratory, Quest Diagnostics, to conduct small-scale test batching through an emergency authorization, the company announced on Saturday, July 18.
Stanford Medicine is also in line for possible emergency-use authorization from the FDA for its specimen pooling, which could markedly increase its daily testing output.
"If we are doing 2,000 tests per day and convert to pooling we can do 20,000 per day," said Dr. James Zehnder, director of clinical pathology at Stanford Medicine.
In specimen pooling, nasal-swab samples taken for the polymerase chain reaction (PCR) diagnostic test, which identifies active virus infections, are still collected in individual vials through swabs used in a patient's upper respiratory system. The specimens are then combined into small batches. A batch that tests negative means all of those patients do not have the virus; if a batch has a positive result, then all of the patients in that pool are retested individually to find out who has an infection, Zehnder said.
"I'm a little concerned there's not a dialogue about reopening schools" using specimen pooling to track infections, he said.
With disease-prevalence rates still low, that could amount to thousands of test kits being freed up. If the FDA approves Stanford's specimen pooling, the university's medical school lab could process up to 10 samples on a test, he said. Stanford could also save an estimated 25% in test and processing costs, according to Zehnder.
Pooling specimens would also help reduce shortages. Various components of the test kits have been in short supply at different times since March: pipette tips, reagent and other chemicals for transporting the tests to labs, Zehnder said. Stanford has been fortunate because it has largely been able to get around the shortages through its innovations. It created its own in-house test, which has FDA emergency-use authorization, and 3D -printed its own nasal swabs, he said. Border towns and areas with fewer resources and many cases have been less fortunate, and specimen pooling could help preserve precious resources and reduce backlogs, he said.
Specimen pooling has been successfully used for more than 70 years. It is currently used at blood banks to screen donations for a variety of viruses.
Donoho said specimen pooling was used in the early days of the HIV/AIDS epidemic, when antibody tests for the virus were restricted to a limited number of kits.
"There are many, many examples. It's proven itself over a lot of years. Mass testing gives us a safe path out of lockdown, and just by using math, we can do that today," he said.
Using specimen pooling, 100 people can be accurately checked for the virus using 10 test kits, Donoho said. Researchers are examining ways to push the frontier further using math and statistics, squeezing the largest number of test samples into the fewest number of tests possible, he said.
More advanced research is also looking at ways to pinpoint people in group testing who test positive so they won't have to be retested individually, he said. Each patient's specimen is divided into a number of subsamples and applied to tests against different sets of patients. The results can be laid out in a table to identify individuals who are likely testing positive.
"We create a kind of a Sudoku puzzle and solve it using math smarts," to find out which people among a group have the virus, he said.
Specimen pooling may have some limitations, including lower test sensitivity if infection prevalence in a population is high, Zehnder and Donoho said. COVID-19 also has a five- to 14-day incubation period, so someone might test negative if they are still at the very early stages of the disease.
Quest's FDA-authorized testing guidelines note that specimens with low low amounts of the virus also may not be detected in sample pools due to the decreased sensitivity of pooled testing. Donoho considers that's less of a problem with SARS-CoV-2, the virus that causes COVID-19.
"The reason is that the RT-qPCR (real-time polymerase chain reaction) test can detect as little as 10 virus particles in a sample," he said.
A dilution of virus particles on a pooled test due to having more than one person testing negative wouldn't have much effect on the test sensitivity because when sick, a person will shed millions of particles, he said.
"It's been proven by several teams that moderate levels of pooling are not a problem," he said.
Quest's study also found no statistically relevant test-sensitivity loss. None of its 3,091 specimens incorrectly tested negative from a population with a COVID-19 community prevalence rate of 1% to 10%, the company said.
Specimen pooling could also make another goal possible: testing people frequently and repeatedly. Continuous testing would help catch the virus in its early stages even after someone initially tests negative, Donoho said.
Stanford's Dr. Benjamin Pinsky, associate professor of pathology and infectious diseases at the School of Medicine, used pooled specimens earlier this year to track the early prevalence of the disease.
Other parts of the country are already conducting pooled specimen testing, such as the state of Nebraska and University of Nebraska system, he said. Cornell University in Ithaca, New York, which will fully reopen this fall, will also pool tests. The university is enforcing a strict regimen of frequent and repeated testing to catch cases, he noted.
"With (repeated) group testing, the benefit is you catch it so early that people are not very infectious," Pinsky said.
Frequent, pooled-specimen testing "makes good sense for universities in college towns," Pinsky said, and he hopes Stanford can use similar testing on its campus, he said.
Dr. Jay G. Wohlgemuth, senior vice president and chief medical officer at Quest Diagnostics, was cautiously optimistic in a July 18 company statement that specimen pooling will help improve testing result times, with a caveat. Just five days prior, July 13, Quest had issued a statement that soaring demand for COVID-19 tests is slowing turnaround times to report results. On July 20, the company said testing was still strained by the demand.
"Pooling will help expand testing capacity but it is not a magic bullet, and testing times will continue to be strained as long as soaring COVID-19 test demand outpaces capacity. Each of us can practice behaviors that will reduce COVID-19 infections in our communities, so our national healthcare system can better respond to this crisis," he said.
Find comprehensive coverage on the Midpeninsula's response to the new coronavirus by Palo Alto Online, the Mountain View Voice and the Almanac here.
on Jul 22, 2020 at 1:13 pm
on Jul 22, 2020 at 1:13 pm
“If that's the case, then the U.S. has a long way to go. Currently, only 7% percent of the population is being tested every day…”
Unfortunately, that’s not what the linked article from SIAM News actually says. Instead, it states that the 7% / day figure is what must be achieved to attain the scientists’ ideal of testing everyone every 2 weeks.
To understand the problem posed by reaching this 7% daily goal, consider that with our county population of almost 2 million, it would mean taking about 140,000 swab tests each day. That will require moving 140,000 people each day to a testing station. Does this sound practical?