Stochastic sampling effects favor manual over digital contact tracing

Mancastroppa, Marco; Castellano, Claudio; Vezzani, Alessandro; Burioni, Raffaella

Isolation of symptomatic individuals, together with tracing and testing of their nonsymptomatic contacts, is a fundamental strategy for mitigating the current COVID-19 pandemic before pharmaceutical interventions become available. The breaking of contagion chains relies on two main alternative strategies: manual reconstruction of contacts based on interviews or a digital (app-based) privacy-preserving contact tracing protocol. We compare in the same framework the effectiveness of the two strategies within the activity-driven model, a general empirically validated framework for network dynamics. Using model parameters tailored to describe SARS-CoV-2 diffusion, we show that even when the probability for a contact to be traced is the same, manual contact tracing robustly performs better than the digital protocol in increasing the epidemic threshold, limiting the height of the epidemic peaks and reducing the number of isolated individuals. This remains true even taking into account the intrinsic delay and limited scalability of the manual procedure. This result is explained in terms of the stochastic sampling occurring during the case-by-case manual reconstruction of contacts of infected individuals, contrasted with the intrinsically prearranged nature of digital tracing, determined by the decision to adopt the app or not of each individual. The better performance of manual tracing is enhanced by the heterogeneous features of agent behavior: a superspreader not adopting the app is completely invisible to digital contact tracing, while she can be traced manually, due to her multiple contacts. Our results indicate a careful integration of the two intrinsically different protocols as key to optimal mitigation strategies.