Federal health officials are hoping to stretch the supplies used to test for the coronavirus by combining samples from a number of people and running a single test. Chinese health officials used that strategy to rapidly test large populations in Wuhan and Beijing.
The technique, called pooled testing, won't resolve the testing bottlenecks in the United States. But it could help.
The idea is simple. Instead of running a coronavirus test on every specimen that arrives in a lab, take a sample of that specimen and combine it with samples from other specimens. Then run a single test on that pooled sample.
If it comes back negative, you can assume that all of the original samples are negative. A single test has done the work of five or 10, compared with testing the samples individually.
If the pooled test comes back positive, it's then necessary to test each of the original specimens individually — and that takes more time and more material. But on the whole, if a lab is testing samples that will be negative at least 90% of the time, the lab comes out ahead using the pooled approach.
"It can be used in any of a number of circumstances, at a community level or even in schools if you wanted to do that," Anthony Fauci said at a recent Senate hearing. Fauci, the nation's top infectious disease doctor, heads the National Institute of Allergy and Infectious Diseases.
This strategy has been used in screening for other viruses, such as HIV, and has already been put to use for COVID-19 in Nebraska.
"The reason why it came up in Nebraska was that we were running out of reagents [testing ingredients] back in mid-March, to be able to run COVID-19 testing in the public health lab," says Peter Iwen, director of the Nebraska Public Health Laboratory.
After convincing himself the technique would work, he got permission from the governor to use this method. He pooled five tests at a time and found that he could run more than twice as many tests with the same materials.
Some health officials have suggested that this technique can vastly increase testing, though unless the virus is rarely present, the savings won't be that substantial.
The Nebraska lab, for example, was able to run more than twice as many tests by pooling five samples. "You can save 50%-60% of the reagents," Iwen says. "You might only be saving 25%-30% of the labor, but you still are saving labor."
There are substantial logistics involved in running these tests. Labs have to keep track of the original specimens and keep track of test results, Iwen says. But large and organized labs can handle these additional steps.
Nebraska actually stopped pooled testing a few weeks ago. An outbreak in meatpacking plants led to a spike in positive test results. It no longer made sense to pool results, since so many of those pools were positive and had to be retested. Iwen says he's developing a method to select groups of samples where positive results will be rare and reserve pooled testing for those cases.
There's another trade-off as well. When labs combine samples, they increase the risk of getting negative results from people who are actually infected.
"That specimen in the pool has been diluted by the rest of the specimens," says Dr. Chris Pilcher, a professor of medicine at the University of California, San Francisco. "So, there's a decrease in the amount of stuff you want to detect in the specimen. And that results in a corresponding decrease in the sensitivity of detection for the test."
Pilcher recently ran an analysis and found it appears to be a manageable problem. He was looking at how quickly viruses arise in samples taken from infected individuals. It turns out that viruses are usually present at high levels, even early in the course of the disease. That means there are relatively few samples with low levels of virus, where concerns about dilution would be greatest.
"You will always lose some proportion of the cases that you could find [if you tested each] individually," he says. But he says the trade-off is worth it if pooling lets you test a lot more people, for example to survey a community or a group of workers or students.
Pilcher explored a system in which samples were pooled in groups of five, and then four or five of those groups were pooled together. If the largest pool tested positive, each of the smaller pools would then be tested. Any that were positive would then require the lab to test all the original specimens.
U.S. Assistant Secretary for Health Adm. Brett Giroir, who is the federal official responsible for wrangling coronavirus testing, says the government is ramping up its efforts to allow pooled testing. He says the initial pooled test could be conducted by a laboratory that hasn't been federally certified, as long as the follow-up testing is performed in a lab that's allowed to run diagnostic tests.
The Food and Drug Administration is developing guidelines for this practice.
Despite the enthusiasm for pooled testing, efficient use of laboratory materials is only one of the challenges in coronavirus testing. Collection swabs and vials can be hard to come by at some labs, and people who want testing can end up in long lines at overworked collection stations. So, pooling alone won't solve the nation's testing woes.
You can contact NPR Science Correspondent Richard Harris at firstname.lastname@example.org.