The application Radar COVID detects twice as many contacts as the manual tracing system
The journal Nature Communications publishes the results of the first scientific study to assess the reliability of the application; one of the participants is the URV professor Àlex Arenas
The application Radar COVID detects twice as many close contacts of people infected with the virus SARS-Cov2 as the manual tracing system. This is the conclusion of the first scientific study that was carried out to assess the application in a trial carried out last summer on the island of La Gomera in the Canary Islands. The following researchers were involved in the project; Àlex Arenas, professor from the Department of Computer Engineering and Mathematics; Lucas Lacasa, from the Queen Mary University, London; and Pablo Rodríguez, from the Association of Computing Machinery, United States. The results have been published in the scientific journal Nature Communications.
The aim of the study was to check the technical and epidemiological effectiveness of the digital tracing of contacts through the application RadarCOVID at the request of the Secretary of State for Digitalisation and Artificial Intelligence of the Spanish government. To this end, a pilot trial was set up on La Gomera on 22 July. The research team simulated a series of COVID infections in the island’s capital, San Sebastián de La Gomera, to verify how accurately the application can detect close contacts and, therefore, contain coronavirus outbreaks, which in this simulated experiment affected 10% of those who had downloaded the application. A publicity campaign was designed to encourage all visitors to the island to download the application. About 33% of the people (more than 3,000 downloads) installed it on their mobiles and 6.3 close contacts were detected per infected individual, a figure that is almost twice the average number detected on the island (3.5) using only manual tracing.
“The number of close contacts that was detected was approximately the same as would be expected in a real situation bearing in mind existing data,” says the researcher Àlex Arenas, who adds that “the percentage of downloads was above the threshold required for the application to be efficient” (approximately 20% according to researchers).
Contacts can be digitally traced by using mobile phone applications that notify users of any recent contacts who have recently been diagnosed as positive for COVID-19. This system has been introduced in countries all over the world to supplement the tracing of manual contacts but this is the first empirical test in real outbreaks.
Despite their usefulness in times of pandemic, this sort of application has been looked on with reluctance by part of the population. One reason for this is the possible danger of detecting a large number of false close contacts that may put the health system under unnecessary pressure. “We have seen that the application can accurately check distances between contacts and we know, for example, that when two people are separated by walls the intensity of the Bluetooth technology decreases,” Arenas explains. Another reason is the privacy of personal data. In this regard, he points out that the app is “totally private” and the user receives only the warning that there has been a contact but with no other information about the identity of that contact. “If you have been diagnosed as positive by a PCR, your health system generates a unique random number that you can voluntarily enter into your app and all the close contacts detected will be sent a notification. This process is totally anonymous and the system is very safe,” he claims.
Even so, the researchers warn that the success of the application depends on the governments, who are responsible for setting up national advertising campaigns to encourage people to download and use the application.
Related Journal Article