Group testing
Researchers have developed a group method of testing COVID-19
Scientists Mario Guarracino from Laboratories of algorithms and technologies for analyzing network structures HSE in Nizhny Novgorod, Julius Zhilinskas and Algirdas Lancinskas from Vilnius University have proposed a new way of testing COVID-19. This is a group method that allows you to get results up to 13 times faster compared to individual testing of each sample. An article with the results of the study was published in the journal Scientific Reports.
The COVID-19 pandemic has already affected millions of people from more than 200 countries. The rapid expansion of the virus has shown how quickly such diseases can spread in today's globalized world. At the beginning of the epidemic, when little was known about the virus and vaccines had not yet been developed, it was possible to mitigate the scale of its spread only by limiting the mobility of people. Almost all people around the world have gone through quarantine and isolation in one form or another. However, restrictive measures can be made less stringent and at the same time more effective if large groups of people are tested quickly, according to the authors of the article "Pooled testing with replication as a mass testing strategy for the COVID-19 pandemics".
Mario Guarracino © Higher School of Economics.
Existing tests for COVID-19 use patients' RNA obtained from oropharyngeal and nasopharyngeal samples, which is checked by PCR for the presence of special RNA strands confirming infection with COVID-19. The speed of such testing is limited by the availability of reagents, laboratories and qualified laboratory technicians.
To speed up testing and increase its coverage, the group testing method is used, in which samples are divided into groups and tested not individually, but jointly. Samples from groups that have shown a positive result are then checked individually to identify specific people. This approach makes it possible to reduce the required number of tests by 2 or more times (depending on the prevalence of the disease) compared to individual testing of each sample.
For example, scientists cite the following calculations: to test 96 people with the ability to include up to 12 samples in one group, 96 tests will be needed for individual testing. For group testing, you will need 8 tests for 8 groups of 12 samples, and then, in case of a positive result of one group, another 12 tests for individual testing. In the case when 2 or 3 groups showed a positive result, it will be necessary to do 24 or 36 tests, respectively. In total, with the first 8 tests, this will result in a reduction of tests from 2 to 5 times compared to individual testing.
The researchers believe that the number of tests can be further reduced by selecting the optimal size of the groups, taking into account the total number of samples and the predicted number of infected. With an increase in the number of infected, the possibility of saving tests decreases, but it still remains at the level of 40% with a morbidity of 100 people per 1000 tested and at the level of 18% with a morbidity of 200 people per 1000 tested.
"Even with a high incidence at the level of 10-20 percent of all tested, the use of OptReplica will be more effective than individual testing," says Mario Guarracino, chief researcher at the Laboratory of Algorithms and Technologies for Analyzing Network Structures.
There are ways to optimize group testing, for example, selecting the optimal group size based on the total number of samples and the estimated incidence rate, or the binary division method, in which the group that showed a positive result is divided in half and tested again until specific positive samples are identified. The second method, however, is time-consuming, which greatly reduces its attractiveness in a pandemic.
In addition, to optimize group testing, the method of replication of groups with transposition is used, when after splitting samples into groups, researchers form additional test groups from the same samples and test them together with the main ones. This makes it possible to further reduce the number of tests, and at low levels of morbidity – to identify positive samples in one stage, which significantly speeds up testing.
Fig. 1. An example of the breakdown of samples into groups when using the method of replication of groups with transposition.
However, this method does not allow experimenting with group sizes to identify the optimal group size under existing conditions. Scientists from the HSE and Vilnius University have proposed the OptReplica technology, which consists in using a more complex algorithm for distributing samples into main and test groups and allows reducing the number of test groups. In addition, the algorithm helps to calculate the optimal group size for the available number of samples and the estimated incidence rate.
Fig. 2. An example of the breakdown of samples into groups using the OptReplica method.
The authors of the technology conducted experimental studies in samples of 96 and 384 samples, 100 randomized trials for each sample size, and compared the effectiveness of replication with transposition and the effectiveness of using OptReplica for different levels of morbidity. Studies have shown that when selecting the optimal size of groups, the use of OptReplica was more effective than the use of replication with transposition. With a low incidence rate, the use of OptReplica made it possible to reduce the number of tests up to 13 times compared with individual testing of the same number of people.
Fig. 3. Reduction of the number of tests when using both the transposition replication method (dotted line) and OptReplica with the optimal group size (solid lines) depending on the level of morbidity when testing a sample of 96 samples (blue color) and 384 samples (red color).
The authors of the new technology suggest using it to test people without symptoms of COVID-19 in regions with presumably low morbidity, where it will allow identifying patients as quickly as possible and with a minimum number of tests and timely applying quarantine measures to them to prevent the spread of the disease. In regions with a morbidity rate of more than 50 people per 1000 tested, the authors recommend using other methods of group replication or testing without replication.
Portal "Eternal youth" http://vechnayamolodost.ru