A strategy to carry out COVID-19 tests in companies or institutions, which optimizes resources and minimizes the presence of people who could be infected, was developed by astrophysicist Xavier Hernández Doring, a researcher at the UNAM Institute of Astronomy. It consists of a novel algorithm that works to identify asymptomatic and presymptomatic patients.
This sampling scheme makes it possible to select optimally and efficiently each day those who apply tests to detect the presence of the virus. SARS-CoV-2 in workplaces, which are part of the standard and mandatory protective measures widely implemented in response to the pandemic.
What does it work for?
Hernández Doring is a theoretical astrophysicist who has been working for some time on problems of gravity, astrophysical jets, galactic dynamics, and statistics of star populations, as well as gamma-ray bursts or galaxies.
However, at the request of his collaborator Sergio Valentinotti Marelli, director of Life Sciences at the Liomont Laboratory and member of the Board of Directors of the Board of Trustees of the Faculty of Chemistry at UNAM, he faced the challenge of designing a plan that would minimize the amount of sick people they have in the plant, reduce infections and not delay production.
“He is in charge of packaging the AstraZeneca vaccines that are being made in collaboration with Argentina, here in Mexico, and he had this specific, real and pressing problem because they cannot work from home. There are machines that need to be operated and the need to get the vaccines out and have people working.”
All employees cannot be given a daily PCR test because there is a test limit that can be processed in a population. That is, if there are a thousand people and they want to be evaluated, at what time do they make the vaccines?
In addition to that, costs, processing capabilities, limited availability, and often cumbersome test requirements must be considered, which dictates seeking the most efficient use of such an important and often scarce resource.
For this reason, the astrophysicist designed a sampling scheme, also published today in the PLOS ONE magazine, where he tests its efficiency with a series of computer simulations to obtain results of the effect it can have. It has been shown to reduce—by as much as a factor of two, even three—the number of presymptomatic or asymptomatic infected people present in a work or educational setting, relative to just doing the same number of randomly selected tests.
Parameters and variables considered by the algorithm to identify asymptomatic patients
- The number of people in the population.
- The mean daily probability of infection in the general population.
- duration of the infection period.
- The number of tests performed each day.
- The number of days after a negative test that a person is excluded from the test sample.
- The number of infected days recorded during a 100-day period.
It is a problem that, obviously, is quite general, because as much as you would like to work from home, there are jobs for which people have to go. For example, hospitals, electricity, food processing, oil, and daily tests cannot be carried out on all personnel.
Just as the need arose in the vaccine manufacturing company, there are industries that should not stop even in pandemics, since services in hospitals, security, electricity or food production must continue, which implies that a significant number of people must go to their labor centers.
In the pandemic generated by COVID-19, it has been especially difficult to identify those infected, because several do not present obvious symptoms; in workplaces or meetings it is key to detect them beyond temperature measurements or the presence of mucus or cough.
“It is then very important to identify them and the scheme becomes very important, since it can reduce the number of infected people present by a factor of two or three using a limited number of tests, in addition to identifying them to give them timely care, also reducing the possibility of death. Contrary to the rest of my work as an astrophysicist, this is something that has a very immediate and obvious use”, stated Hernández Doring.
The population in which the efficiency of the scheme was tested is greater than a thousand people. The researcher explained that it was possible to identify a significant number of asymptomatic or presymptomatic women, in figures that coincided with the theoretical simulations presented in the article, for which he estimated that without its application the company could have had more cases of contagion.
He emphasized that the algorithm was specifically designed for this purpose and is not used in astrophysics, so he hoped that it would be possible for more industries or offices where they return to activities to take advantage of it, hence its presentation in a free access peer-reviewed journal.