Also known as batch testing, pool testing combines samples from several people and tests them for the coronavirus all at once, cutting down on the time and supplies required. The protocol was first invented to test for syphilis during World War II and has been used in the past for outbreaks of other sexually transmitted diseases, including HIV.
“If everyone is negative, then you’re done,” Ashish Jha, director of the Harvard Global Health Institute at Harvard’s T.H. Chan School of Public Health, told The Post in May.
If the test detected the presence of the virus, then each person would have to be tested and the results individually analyzed to determine whose sample produced the positive result.
“You can rapidly increase the capacity of testing,” said Benjamin Pinsky, director of the Clinical Virology Laboratory at Stanford University’s School of Medicine. “The trade-off is that there’s reduced sensitivity. It’s kind of a balance.” Samples with low viral loads are more likely to go undetected in a pool, he said.
How many samples are pooled?
Researchers have generally suggested quantities between three and 50. The bigger the pool, the more likely a positive case with a low viral load will be too diluted to trigger detection of the virus. Israeli researchers found that a positive sample was detectable in a pool of 32 and even a pool of 64, although the latter group might require more cycles of amplifying the virus. A large pool that contains a positive case also means more individual tests will have to be run to find the infected person. But researchers are also experimenting with ways to organize the samples that will make it easier to disaggregate a positive case from a pool.
When does pool sampling make sense — and when does it not?
Pool sampling works best in a large population in which the infection rate is believed to be low. If you are working with a small group, pool testing isn’t necessary. If you are working with a large group but suspect many of those people are positive, it’s not helpful to group them together, because many of the samples will have to be retested individually.
Jha said the method could be used, hypothetically, to test children attending a summer camp in a low-transmission area. “You use one test kit, and if it’s negative, then you say, we’ll have the camp. And if it’s positive, we’ll send everyone home and tell parents to figure out who is positive,” he said.
Jha and other experts suggested that other group settings in which regular pool testing might be effective include schools, offices and correctional institutions.
The concern over false negatives would be mitigated if the same pool is tested multiple times, offering more chances to catch the virus.
“It could be very useful in back-to-work or back-to-school programs or large-scale screening because this allows you to test more people and test them more frequently, and that’s the key,” Pinsky said.
Is anyone else doing this?
Countries around the world have been doing pool testing. India’s protocol combines 25 samples at a time for migrant workers and international travelers in quarantine and for surveillance in “green zones” where prevalence is low. Germany has used pool testing in nursing homes and residential care facilities. Singapore has as well.
If it’s so great, why aren’t we doing it already?
In the United States, federal restrictions have limited the availability of batch testing. The Food and Drug Administration considers pooling samples a modification that takes a test out of compliance. One exception is Nebraska, where Gov. Pete Ricketts (R) gave the state public health laboratory emergency authorization to pool tests.
“We were running out of reagents. We were at a point here we were going to have to shut the lab down,” lab director Peter C. Iwen recalled.
The FDA subsequently agreed that the lab could pool up to five tests at a time as long as the overall positivity rate remained below 10 percent. Iwen said he is in conversation with pathologists and the FDA about how to allow national pool testing. Just 10 days ago, the FDA released a template for labs to seek approval of pool testing protocols.
“There’s a huge change going on . . . to allow this to happen,” he said. “This is what we need. Everybody is skittish about doing this because of the regulations.”
Pinksy’s lab at Stanford has also done some pool testing of San Francisco residents who were swabbed for respiratory symptoms from December through February — before a coronavirus test was available. Using groups of 10, they tested 2,888 samples and found only two positive cases, indicating the virus was not widespread in the area at that point.
The lab is now preparing for clinical pool testing, pending approval from the FDA.
Paige Winfield Cunningham and Rosalind S. Helderman contributed to this report.