Fast testing method can ID hot spots
Stanford set to swab pool of 8 people at same time
Stanford Medicine plans to start a promising method of testing for coronavirus that could significantly speed up efforts to identify cases in the Bay Area.
“Pooled testing” — in which a lab processes specimens from multiple people at one time, rather than just one specimen at a time — could potentially be used to conduct frequent, largescale tests at schools, dorms, nursing homes and prisons. That would make testing faster and more efficient, and thus help prevent outbreaks in those settings.
Pooled testing has long been used to test large numbers of blood donations to make sure they are not infected with HIV or hepatitis B and C, and to screen for syphilis in the military.
Stanford will test for the coronavirus in pools
of eight. That means specimens from eight people, who were swabbed individually, will be combined and tested together in the lab. A part of each person’s specimen will be preserved and set aside in case it needs to be retested. If the entire pool tests negative, there is no need to test each specimen separately — saving time and lab resources. If the pool tests positive, only then would each individual specimen need to be tested.
For that reason, experts say pooled testing works best for assessing lowrisk asymptomatic people who are likely to test negative, and not symptomatic people who are more likely to test positive.
Stanford plans to start pooling tests within the next two weeks. It is expected to increase Stanford’s daily lab capacity from 4,500 tests to 20,000 tests.
“It’s going to allow us to test classrooms and schools and dorms and longterm care facilities and prisons, and be able to screen those populations very frequently,” said Dr. Benjamin Pinsky, director of the Clinical Virology Lab at Stanford Medicine. “The test volumes required to do that if we did every person individually is not possible . ... It will allow efficiency at a large scale. Otherwise, we’re not going to be able to (test) those populations the way we need to, to prevent outbreaks.”
Stanford in late June applied for FDA Emergency Use Authorization to do pooled testing, but has not yet received a response from the agency. Stanford expects to get a response this week. The FDA did not immediately respond to questions from The Chronicle about Stanford’s pooled testing. But in a July 8 virtual town hall meeting, FDA officials said labs can move forward with pooling as long as they have validated that their pooling protocols work, notified the agency, and submitted validation data — which Stanford says it has done.
About twothirds of the coronavirus tests processed at Stanford’s lab are from patients of Stanford and ValleyCare Medical Center, in Pleasanton. The remaining third come from patients in the broader Bay Area, including essential workers, health care workers and patients at community health clinics.
Since Northern California schools, nursing homes and other facilities get priority over outofstate facilities to have their tests processed at Stanford’s lab, it is a promising step for the region in helping those local facilities test faster and more efficiently.
Stanford is one of just a small number of labs in the country, and one of the only labs in the Bay Area, to conduct pool testing. Santa Clara County’s Valley Medical Center, which has been conducting the bulk of coronavirus testing in the county, has also been pooling tests since late May, in groups of five.
Others include the Nebraska Public Health Lab, which pooled tests earlier during the pandemic but later stopped, and the University of Florida.
One potential drawback is that if there is a positive case among the pooled specimens, pooling dilutes the concentration of viral RNA and makes it harder to detect the positive. Pinsky said pooling makes testing about 10 times less sensitive. That raises the possibility of false negatives.
False negatives already occur in individual testing, usually because the patient tested too early or too late — when their virus load is too low to be detected by the test — or because the quality of the specimen was not good.
Stanford tested 880 specimens individually, and in pools of eight, and found that pooling detected 70% to 80% of the positive cases that had been detected through individual testing.
Pooled testing was more likely to detect a positive case if that specimen had high levels of the virus, and less likely to detect the positive case if the specimen had low levels of the virus.
Pooling experts say that tradeoff is worth it for testing certain populations where individual testing is not feasible or would take much longer. Even though the proportion of detected cases in pooled testing is slightly lower, the number of tests that can be done through pooling is so much bigger that the process ends up detecting many more cases than what would be detected through individual testing, said Dr. Chris Pilcher, a professor emeritus at UCSF School of Medicine.
“In some circumstances, some loss of sensitivity is acceptable because the real comparison we’re talking about is not between the sensitivity of pooled testing and the sensitivity of individual testing. It’s between the sensitivity of pooled testing and the sensitivity of no testing at all,” said Pilcher, who helped spearhead the nation’s first and largest pooled testing program, for diagnosing HIV, at North Carolina state labs nearly 20 years ago.
Quest Diagnostics, the large national commercial lab that is conducting a huge portion of U.S. coronavirus tests, received FDA Emergency Use Authorization over the weekend to conduct pooled testing in groups of four.