Ariadna Fosch i Muntané
Predicting the effect of contact-tracing apps in epidemic spreading: A multilayer network approach
Numerous countries have relied on contact-tracing (CT) applications as an epidemic control measure against the COVID-19 pandemic. However, limited research has been done to characterise their effectiveness in a dynamic adoption setting.
To this end, we propose the implementation of a multilayer network approach to represent the co-evolution of an epidemic dynamic, modelled using a modified SEIR model, and the CT app adoption process, represented by a threshold dynamic over the same population. The model was initialised to reflect COVID-19 progression and implemented over two populations with different network structures. Two different strategies for the implementation of CT apps were explored, which allowed to characterise the relevance of the time of adoption, number of adopters and level of compliance for the effectiveness of the app.
The results highlighted that an early implementation of the strategy and an adoption by more than 25% of the users is required to ensure a minimal effectiveness (>5% mean peak reduction). Moreover, compliance with the strategy also plays a major role in the app's performance, since even low levels of compliance can result in a moderate effectiveness (>10% mean peak reduction) if high levels of adoption are considered.
The insight obtained from this project was used to study the Spanish CT app in which we identified a bottleneck in the process for reporting an infection. We suggest that the implementation of a more straightforward verification process would increase the compliance with the strategy and cause an increase in its effectiveness.