Real-Time Epidemiological Platform.

Technology to the rescue. For several weeks now, I have been closely following more than 12 global applications designed for COVID-19 pandemic protection. What stands out is that the major debates around privacy and discrimination are largely absent from their development. Even more striking: the technology already within reach could, in theory, effectively halt the pandemic.

My intention is not to survey multiple applications, but to describe key aspects of the contagion-risk calculation protocol developed by Google and Apple. Once I have done that, I want to put forward several ideas I have explored in depth regarding the core functionalities these systems should have at a global level. Let us be clear: this pandemic requires a globally interoperable system, not merely local solutions.

The debate over privacy and discrimination has its nearest roots in the tragic fight against HIV. The discrimination that community continues to face today reminds us how little progress we have made in separating people from their illnesses. Protecting the privacy of diagnoses also serves as a safeguard against the failure to build genuine public awareness and education around health conditions. We should not forget that privacy implications also carry the risk of an irresponsible state weaponizing information about who met whom.

Apple and Google's contact-risk calculation protocol uses four factors: transmission risk, contact duration, days since exposure, and contact distance. The weights assigned to each factor, defined by each country's health authorities, allow every user within the contagion-risk network to know their individual risk score in complete privacy. https://developer.apple.com/documentation/exposurenotification/enexposureconfiguration

The variables work as follows: A. "Transmission risk": a symptom questionnaire assigns values to the probability of infection. B. "Contact duration": a value is assigned based on how long individuals or groups spent together. C. "Days since exposure": how long ago the exposure occurred, and if certain symptoms have not developed, the user is presumed non-infected or asymptomatic. D. "Contact distance": calculated by measuring the distance between devices, inferred from the Bluetooth signal strength at which those devices detected each other.

Taken together, these variables allow users to maintain a contact history stored on their own devices as they go about daily life. If this data were held in a centralized system, the risk of repurposing would be self-evident. This decentralized design is a significant innovation, though it requires a complementary testing infrastructure: a system that triggers alerts whenever a user tests positive and, as the algorithm indicates, notifies the relevant contacts.

What core functionalities should such a system have? Three are essential: 1. Individual risk-level awareness. 2. Real-time data feeding a global dashboard of contagion clusters and risk zones. 3. Streamlined contact-tracing. A concrete example: if a workplace registered elevated collective risk based on the aggregate of individual scores, companies could be empowered (or required) to send flagged employees for testing, directly reducing community transmission.

Looking further ahead, the applications of this data become a gold mine for understanding urban social dynamics. Identifying the variables (voluntarily shared by users) that correlate with higher infection rates raises critical questions. Is overcrowding the driver? Is it transport hubs and mobility corridors? This kind of information is valuable not only at the individual level, protecting those we care about by managing our own risk responsibly, but also at the collective level, as a shared obligation to continuously monitor ourselves.

Multiple countries are already using this type of software. The principal advantage of the Google and Apple API is that it protects privacy and lets users decide what data they share, from location to sociodemographic characteristics. Only through voluntary global cooperation will it be possible to build a platform capable of finding new equilibria in public and social functioning.

Looking beyond COVID-19, a Global Urban Epidemiology System could also protect populations from other conditions and diseases. Influenza remains a significant killer, and having real-time global data could enable prioritized action at a planetary scale. Imagine global tracking of any transmissible infection, with tools that help individuals assess the risk of going somewhere and, conversely, allow those in high-risk corridors to take targeted action to reduce exposure. In the aggregate, this approach meaningfully increases the health prospects of all humanity.

Frequently Asked Questions

How does the Google and Apple exposure notification protocol protect user privacy?

All contact history is stored on the user's own device rather than a centralized server. Risk scores are calculated locally, and users control exactly what data, if any, they choose to share with health authorities or the broader network.

What are the four factors in the Google-Apple contact-risk calculation?

The four factors are transmission risk (derived from a symptom questionnaire), contact duration (time spent together), days since exposure, and contact distance (inferred from Bluetooth signal strength between devices).

What would a global urban epidemiology system enable beyond COVID-19?

A globally interoperable epidemiological platform could provide real-time tracking of influenza, respiratory infections, and other transmissible diseases, enabling planetary-scale prioritized interventions and helping individuals and institutions make informed, data-driven risk decisions.