Stochastic sampling effects favor manual over digital contact tracing

Area of Science:

• Epidemiology
• Network Science
• Computational Biology

Background:

• COVID-19 pandemic mitigation relies on isolating symptomatic individuals and tracing contacts.
• Contagion chain interruption uses manual interviews and digital (app-based) contact tracing.
• Understanding the comparative effectiveness of these strategies is crucial for public health.

Purpose of the Study:

• To compare the effectiveness of manual versus digital contact tracing for mitigating COVID-19.
• To analyze the impact of individual behavior and tracing protocols on contagion dynamics.
• To determine the optimal strategy for realistic hybrid public health interventions.

Main Methods:

• Utilized an activity-driven model, an empirically validated framework for network dynamics.
• Incorporated model parameters specific to SARS-CoV-2 diffusion.
• Compared manual contact reconstruction with digital, privacy-preserving contact tracing protocols.

Main Results:

• Manual tracing demonstrated superior effectiveness compared to digital tracing, even with inherent delays and scalability limitations.
• The stochastic sampling in manual tracing contrasts with the prearranged nature of digital tracing.
• Heterogeneity in agent behavior, particularly superspreaders, significantly favors manual tracing's effectiveness.

Conclusions:

• Manual contact tracing is a more robust strategy for controlling COVID-19 transmission than digital methods alone.
• Digital tracing's effectiveness is limited by individual adoption rates and inability to capture superspreaders.
• Hybrid protocols incorporating manual tracing are dominant for realistic pandemic control.