Israel Introduces Pooling Method to Increase Coronavirus Testing Capacity

The combinatorial testing strategy conserves resources and can allow Israel to perform hundreds of thousands of COVID tests a day

Send in e-mailSend in e-mail
A testing lab for the coronavirus in Israel, April 2020.
A testing lab for the coronavirus in Israel, April 2020.Credit: Moti Milrod

Clinics belonging to health maintenance organizations have started using pooling methods to make coronavirus testing more efficient, three months after getting approval for doing so by the Health Ministry.

Together with an expansion of testing using private companies, Israel will be able to test hundreds of thousands of samples a day.

Pooling was first used during World War II to identify soldiers with syphilis. The basic idea is to pool several samples and test the combined one. If it’s negative, all the people tested are free of the disease. Only if it’s positive are the individuals contributing to the pool re-tested to identify the specific person with the disease. This speeds up testing and saves on raw materials.

In addition to simple pooling one can use combinatorial methods to make the test even more efficient. One can, for example, test 100 samples using 20 tests. One can imagine the samples as lying on a 10x10 grid. First you pool them horizontally and test them, then you pool them vertically and re-test. Crossing the results can detect a positive sample.

China has already used pooling methods for rapidly testing millions of people, as part of its effort to block the spread of the virus. Slovakia has also announced a campaign to test all its 5.5 million citizens aged 10 or above within one week. Over the last weekend, two thirds of its residents were tested, receiving results within 30 minutes. The campaign will be completed this weekend. With a capacity to test hundreds of thousands of samples a day, Israel can also embark on such a campaign in an attempt to overcome the epidemic.

The pooling of samples is an issue receiving unprecedented attention since the outbreak of the epidemic, after authorities around the world realized that testing for the virus must be made more efficient. Two methods that were approved last August were validated this week at a Clalit HMO clinic, at the Soroka Medical Center in Be’er Sheva, and at a Kupat Holim Meuhedet clinic. These methods were subsequently put in place, with the first samples tested as part of routine operations at Soroka.

The first method was developed by Prof. Angel Porgador and Prof. Tomer Hertz from Ben-Gurion University and the National Institute of Biotechnology in the Negev, along with Prof. Yonat Shemer-Avni, head of the Virological Clinic at Soroka, and Prof. Noam Shental from the Open University. The second pooling method was developed under the leadership of entrepreneurs Ruth Polachek and Tsvika Vagman.

A drive-through coronavirus testing facility in Jerusalem, September 2020. Credit: Ohad Zwigenberg

Shental, Porgador and Hertz’s method is much more complicated. If the method noted above involved two dimensions, they use 48 dimensions. They take 384 samples and divide them into 48 pools, each one with material from the original samples. The calculation and application of such a complex division cannot be done in a lab by hand or with pipettes. Thus, the new setup includes using a robot for performing the pooling.

“It’s exciting to see what began as a phone conversation between Noam and myself last March becoming a method used by the state,” says Hertz, adding that the Health Ministry has acquired 12 robots for pooling samples. He says that distributing the new equipment and methods will be done gradually, based on the success of their implementation in clinical labs.

One of the limitations of pooling is that their efficiency declines with an increase of positive carriers in the population. “Now, when 2 percent of people tested are positive, the method is optimal. But we know that rates can go up significantly,” says Hertz. Polachek and Vagman’s algorithm allows for adapting pooling to the changing incidence of positive carriers in the population. Hertz says that they are now working with software company Neura for embedding the algorithm into artificial intelligence. The algorithm is based on individual medical data of people being tested, used anonymously, comparing them to patterns of movement in any area. This is being examined with a database used by the Clalit HMO, and will enable predictions of the probability that people being tested are positive. This will allow pooling to continue when infection rates increase.

Hertz adds that they’re now working on using this method in the commercial lab at Ilex, a company situated near the airport. This lab can handle 10,000 samples a day. Integrating pooling will enable a significant increase in its capabilities. “The application of pooling methods, probably the most advanced in the world, is a statement of intent by the state” says Hertz. “A clever use of testing capabilities will allow the epidemic to be overcome.”