Samples scraped from the back of patients’ noses and throats can now be pooled and analyzed for clumps of the coronavirus at one of the country’s largest commercial laboratories. Quest Diagnostics, which handles tens of thousands of coronavirus tests each day, was cleared for the method, called the combined test, from the Food and Drug Administration over the weekend.
The strategy helps conserve resources by running multiple tests using the materials generally required for one. It could also help Quest avoid delays and delay testing more quickly. Right now, the demand for testing is so high that it takes Quest Labs an average of seven days to get results for patients.
However, the strategy cannot remove all of the testing hurdles, and it may not work as well in the areas where it is most needed: outbreak hotspots, where testing resources are lengthened the most. In those settings, where a high percentage of tests are positive, groups do not end up saving materials; too many are positive, which means more testing.
“When the positivity rate is high, clustering is not very effective, because it’s dividing too many clusters, and it’s just not worth it,” says Peter Iwen, director of the Nebraska Public Health Laboratory.
The Quest method will combine samples taken from four people. If the multiple-sample test is negative, all four are considered negative. If positive, each sample is retested on its own to determine who has the virus. The company said in a press release that it will begin implementing cluster testing at its laboratories in Chantilly, Virginia and Marlborough, Massachusetts, later in the week, before expanding to other locations. Quest did not respond to a request for comment.
Experts have been interested in clustered testing for months. The method is already commonly used to detect sexually transmitted infections in people and also to detect diseases in donated blood. Deborah Brix, a member of the White House coronavirus task force, said at a conference of the American Society for Microbiology last month that the strategy could dramatically increase the number of tests conducted in the United States each day.
In March, Iwen began exploring pooled tests at the Nebraska Public Health Laboratory. You were running out of the chemicals needed to run each test and you knew you could keep them by pooling your samples. She did some experiments and found that she could do it successfully. “It works very well, and I can save 60 to 70 percent of our reagents,” she says. He obtained approval to advance from the Governor of Nebraska under his state of emergency.
“We started clustering and we were able to keep the lab running,” says Iwen.
The shortage of testing materials is just the kind of hurdle that a combined testing strategy can help overcome, says Nikhil Gopalkrishnan, a postdoctoral fellow at the Wyss Institute for Biological Engineering at Harvard. But a shortage of supplies isn’t the only problem holding up testing. “Depending on what the real problem is, clustering may or may not have a real impact,” he says. “If it turns out they are understaffed, and they just don’t have enough people, and the limiting factor is that someone has to unscrew all the tubes, the bundle won’t help them there.”
Clustering does not help either once the infection rate in a community begins to increase. Iwen had to stop using the pooled tests in the Nebraska lab a few weeks ago because the percentage of tests that tested positive was too high, above 10 percent. In any random group of four or five tests, there was a high probability that one was positive. If the positive rate is so high, pooling does not tend to save a significant amount of chemicals or time.
Quest’s new instructions for combination testing say that if 13 percent of tests done in a lab yield positive results, the lab should reduce the group size to three. If 25 percent of laboratory tests give positive results, they should not do pooled tests. Virginia and Massachusetts, home to the first Quest labs to use pooled tests, the positivity rate is approximately 6 and 2 percent, respectively. In Arizona, one of the most affected states, it is around 24 percent.
Clustered tests can be made to work in an area with a high percentage of positive tests, says Iwen. If they have information about patients (if they have symptoms, where they live, where they work), they can classify them as high or low risk. “You could group those that you think are considered low risk. You could do it anywhere, even if your overall rate is very, very high, “she says.
That only works if labs can get good demographic information on the samples they are testing. But often, labs collect samples and have no idea where they come from, Iwen says.
Despite the limitations, Gopalkrishnan says Quest authorization is a good first step toward more generalized grouped testing. You would like to see larger pool sizes, particularly in areas where there is not much COVID-19 spread. Grouping pools of up to dozens of samples, or up to 100, could help labs move through samples even faster. “I think you will see more bundling authorizations,” he says.
Iwen believes that implementing grouped tests correctly can help tests move faster. “This is a process that must be considered if we are going to do the tests that have been asked of the labs,” he says. “Response time is important, if it has an impact on that, it’s worth it.”
