[论文解读] How to Return to Normalcy: Fast and Comprehensive Contact Tracing of COVID-19 through Proximity Sensing Using Mobile Devices

Ye Xia, Gwendolyn K. Lee|arXiv (Cornell University)|Apr 27, 2020

COVID-19 Digital Contact Tracing参考文献 7被引用 30

一句话总结

该论文主张基于近距离感知的、基于设备的接触追踪策略，以便快速而全面地识别并隔离感染，极高的采用率可能使社会恢复正常。

ABSTRACT

We outline a contact-tracing strategy based on proximity sensing using mobile devices. We discuss what an ideal system should look like and what it can do. We show that, when adopted sufficiently broadly, such a contact-tracing strategy can bring COVID-19 under complete control, end the need of social distancing, and return the society to full normalcy. We also review some of the challenges faced by the current generation of proximity-sensing technologies, including Bluetooth Low Energy used by phones, and consider both interim and longer-term solutions. Our main contribution is that we reason through why such a contact-tracing strategy is likely to achieve the stated goal of returning to full normalcy. Using probabilistic models, we show that universal adoption is not necessary to achieve the stated goal, thus there is some room for exceptions; however, the adoption rate needs to be very high, e.g., above $95\%$ depending on the disease parameters. With more vigilance in disease surveillance to detect mild cases earlier, the number may be brought down to about $90\%$. The results call for deployment effort to be led by public authorities at the state or federal level so that the required adoption rate can be reached and the tracing coverage is wide enough to be relevant for disease control.

研究动机与目标

• 提出一种基于移动设备的接触追踪策略（接触记录器），利用近距离感应来迅速检测密切接触。
• 论证普遍采用并非严格必要，但需要很高的采用率以确保对传播的控制。
• 提供概率分析以推导采用率阈值，并讨论实际部署与隐私考量。
• 评估用于追踪的近距离感应的设备需求，以及临时和长期的解决方案
• 强调公共当局为推动广泛采用应采取的政策含义。

提出的方法

• 在追踪和隔离下用有效再生数 Re 建模传播。
• 定义并分析 X_i 对有效感染数 Chi 的贡献，其中 Re = R0 * P(X_i=1)。
• 通过 定理1 证明，存在有限的采用率 p0，使得对所有 p>p0 Re<1，即使采用并非普遍。
• 在症状报告概率和簇动力学下推导最小采用率 p* 的下界和上界。
• 给出一个基于泊松的直接感染示例以说明 pi0 和 pi1 值（Table II）。
• 讨论实际设备要求以及隐私保护考量，并与现实部署（TraceTogether、Apple/Google 的努力）进行比较。

实验结果

研究问题

• RQ1 Is universal adoption strictly necessary to keep infections under control with proximity-based tracing?
• RQ2 What adoption rate p* is required for Re<1 under varying R0 and symptom probabilities?
• RQ3 How does the tracing strategy perform in terms of complete traceability of infection clusters and timeliness of detection?
• RQ4 What are interim vs. long-term hardware/software design considerations for contact recorders?
• RQ5 What are policy implications and privacy considerations to enable broad adoption?

主要发现

R0	pi0	pi1	1-epsilon
3	0.0595	0.9405	0.6667
4	0.0198	0.9802	0.75
5	0.0070	0.9930	0.8
6	0.0025	0.9975	0.8333

• Universal adoption yields complete traceability of infection clusters, enabling effective containment.
• There exists a finite adoption threshold p0 such that Re<1 for all p>p0, indicating non-universal adoption can still control outbreaks under certain conditions.
• For COVID-19 parameter ranges, the required adoption rate is very high (e.g., above ~95%), with room for slight exemptions.
• Table II provides pi0 and pi1 values under Poisson offspring with R0=3,4,5,6, showing high likelihood of cluster extinction under tracing (pi1 close to 0.94–0.997).
• The strategy emphasizes real-time, wide sharing of contact information within clusters to minimize delay in traceback and isolation.
• The work discusses privacy-preserving approaches and advocates public authority-led deployment to achieve broad adoption.