Network-based proactive contact tracing: A pre-emptive, degree-based alerting framework for privacy-preserving...

Area of Science:

• Epidemiology
• Network Science
• Public Health Technology

Background:

• Current COVID-19 exposure notification apps utilize binary contact tracing (BCT), which issues uniform quarantine orders based on a fixed risk threshold.
• BCT's limitations include delayed alerts and broad isolation, potentially missing early transmission while causing unnecessary disruption.

Purpose of the Study:

• To introduce Network-based Proactive Contact Tracing (NPCT), a novel, privacy-preserving, and decentralized intervention scheme.
• To evaluate NPCT's effectiveness in mitigating epidemics compared to traditional BCT.

Main Methods:

• NPCT utilizes users' Bluetooth contact history to generate individual risk scores.
• A dynamic, epidemic-aware threshold adjusts intervention levels, triggering graded "reduce contacts by X%" prompts instead of all-or-nothing quarantines.
• Simulations were conducted on four synthetic and empirical temporal networks.

Main Results:

• NPCT demonstrated the ability to reduce epidemic peaks by approximately 40%.
• The intervention suppressed only 20% of contacts, indicating a more targeted approach.
• Intervention burden was concentrated on high-risk individuals, with stable performance across different network types and compliance levels.

Conclusions:

• NPCT presents a practical and effective upgrade for existing BCT applications.
• The scheme balances epidemic control with enhanced privacy protection and reduced social costs.
• NPCT offers a more nuanced and efficient approach to managing infectious disease outbreaks through digital contact tracing.