Digital Contact Tracing: an Erroneous Tracking Device or the Savior of Humanity?
In early 2020, 5,000 people were infected with Sars-CoV2 in South Korea, all of which rooted from one single person (Kojaku et al., 2021). At the time, the number accounted for more than half of the total national cases (Kojaku et al., 2021). Despite the high number of cases, they were all identified within months, and this success is being attributed to digital contact tracing (DCT) (Kojaku et al., 2021). Contact tracing identifies individuals who have been in contact, meaning who might have infected the disease, from an identified patient. Contact tracing was done using the manual method in the past, where trained personnel had to physically track down the contacts. Today, especially with the current pandemic, most countries have switched to DCT as it is much faster and effective than the manual method. In fact, many countries have their own DCT apps: TraceTogether of Singapore, SwissCovid of Switzerland, Radar Covid of Spain, etc. (Lewis, 2021). Yet, some experts have cautioned about the social implications of DCT, such as the digital divide between different demographics and privacy issues.
Most DCT apps make use of Bluetooth on smartphones, while others may also utilize GPS (The Tech, 2020). Bluetooth involves the exchange of codes between phones (The Tech, 2020). This code contains no information about the user’s personal information or location (The Tech, 2020). When a person tests positive for COVID-19, the codes are used to send notifications to the cell phones that have shared the codes with the patient’s mobile phone, which alerts the cell phone owners to take the necessary COVID-19 precaution measures, such as getting tested or being quarantined (The Tech, 2020).
Many contact tracing apps prompt smartphones to share codes only under specific conditions to minimize error. To illustrate, let’s say we have person X, Y, Z. Person X was chatting with Y for over an hour. In contrast, X was standing next to Z at a bus stop only for about 30 seconds. The next day, X finds out that they have COVID-19. Evidently, it is more likely for X to have transmitted the disease to Y than Z. Therefore, it would be inefficient to alert Z to get tested or quarantined, wasting resources and inconveniencing Z. Contact tracing apps resolve problems like this by sharing codes only when the likelihood of disease transmittance is high (Lewis, 2021). For example, the Google/Apple Exposure Notification (GAEN) system is the main Bluetooth technology being used in DCT apps (Lewis, 2021). The GAEN system prompts two smartphones to share a code only if they have been closer than 2 meters for at least 15 minutes, aligning with the standard time and distance for Sars-CoV2 transmission (Lewis, 2021).
Nonetheless, DCT has its limitations. For one, sometimes Bluetooth signals can penetrate physical barriers, such as walls, which would have obstructed infection transmission (The Tech, 2020). Ultimately, DCT is only as effective as the number of people who are using the app (Soltani et al., 2020). This phenomenon is based on the MetCalf’s law, which states that the effectiveness of an online network is equal to the square of the total individuals connected through that network (Metcalfe’s law, 2008, 14). For example, in the US, 81% of the population has a cell phone (Soltani et al., 2020). Assuming that the entire 81% downloaded the app, 65% of the total transmission events would be detected by DCT (Soltani et al., 2020). The problem is further exacerbated by the fact even amongst the people who have smartphones, not all of them might’ve installed the app, such as seniors who might be unfamiliar with downloading apps (Soltani et al., 2020).
Despite concerns over its effectiveness, DCT is incomparably faster than the manual method, making it invaluable for controlling the pandemic. A 2021 study on the Radar Covid app in Spain demonstrated that the manual method was able to detect only half the number of people exposed to the virus compared to DCT (Rodríguez et al., 2021, 1). A team of researchers at the University of Oxford analyzed DCT in Britain and presented that “every 1% increase in app users — above a minimum of 15% — reduces the number of infections by 0.8–2.3%” (Lewis, 2021). Quick identification and isolation of affected or potentially affected people are the key to mitigating infectious diseases, cutting off the transmission cascade earlier on.
Consequently, various initiatives are taking place to increase the reliability of DCT. One focal area of research is on improving the digital app’s capability to assess the user’s risk of infection (Lewis, 2021). For instance, was the user together with the COVID-19 patient indoors or outdoors? Does the user have a biological predisposition which makes them more prone to transmitting or getting infected? All of these factors influence a person’s risk of infection, and thus are useful statistics that could be incorporated into the digital app. While collecting more information may increase accuracy, it may discourage the public from using the app in fear of their privacy being invaded; Singapore’s TraceTogether app was criticized as the data it recorded was permitted to be used for police investigations (Lewis, 2021). Conversely, not enough data would increase DCT error rate, which would also decrease public trust in DCT.
The solution seems to be increasing the accuracy of contact tracing measures, without compromising the privacy of the users to a significant degree. Viktor von Wyl, an epidemiologist who studied Switzerland’s SwissCovid app, puts it best: “It’s a fine balance between adding more information, or getting more information out of it, but then possibly losing more users because fears of privacy have increased” (Lewis, 2021). What is considered as enough information is a topic beyond the scope of this article, and may be left unsettled for a long time. One thing is clear: we are all hoping for a day to take that mask off, and a compromise between protection and privacy seems inevitable.
References
Kojaku, S., Hébert-Dufresne, L., Mones, E., Lehmann, S., & Ahn, Y.-Y. (2021, February 25). The effectiveness of backward contact tracing in networks. Nature Physics. https://doi.org/10.1038/s41567-021-01187-2
Lewis, D. (2021, February 26). Contact-tracing apps help reduce COVID infections, data suggest. Nature. https://www.nature.com/articles/d41586-021-00451-y
Metcalfe’s law, Web 2.0, and the Semantic Web. (2008, February). Journal of Web Semantics, 6(1), 14–20. https://doi.org/10.1016/j.websem.2007.11.008
Rodríguez, P., Graña, S., Alvarez-León, E. E., Battaglini, M., Darias, F. J., Hernán, M. A., López, R., Llaneza, P., Martín, M. C., Ramirez-Rubio, O., Romaní, A., Suárez-Rodríguez, B., Sánchez-Monedero, J., Arenas, A., & Lacasa, L. (2021, January 26). A population-based controlled experiment assessing the epidemiological impact of digital contact tracing. Nature communications, 12(587), 1–6. https://doi.org/10.1038/s41467-020-20817-6
Soltani, A., Calo, R., & Bergstrom, C. (2020, April 27). Contact-tracing apps are not a solution to the COVID-19 crisis. Brookings. Retrieved March 23, 2021, from https://www.brookings.edu/techstream/inaccurate-and-insecure-why-contact-tracing-apps-could-be-a-disaster/
The Tech Behind COVID-19 Contact Tracing. (2020, August 4). WatchBlog: Official Blog of the U.S. Government Accountability Office. Retrieved March 23, 2021, from https://blog.gao.gov/2020/08/04/the-tech-behind-covid-19-contact-tracing/#:~:text=Contact%20tracing%20apps%20use%20digital,anonymous%20codes%20shared%20between%20phones
