Abstract

Using Google-Apple Exposure Notification (GAEN) and Exposure Notification Private Analytics (ENPA) to Support COVID-19 Pandemic-response— 13 U.S. Jurisdictions, July 2021–April 2022

Allison Hankus, MPH¹, Ernist Kim¹, Daniel Eisert¹, Christine Cho¹, Sarah Gran², Mariana Raykova, PhD³, Rakhee Palekar, MD, MPH¹ (1)MITRE, (2)Internet Security Research Group, (3)Google

APHA 2022 Annual Meeting and Expo

BACKGROUND: Infectious disease contact tracing is time and labor intensive for public health authorities (PHAs) and relies on the infectious person attempting to recall their contacts. During the COVID-19 pandemic, Apple and Google, along with their partners, developed a protocol that leverages smartphone Bluetooth capabilities, to perform digital contact tracing, in a privacy-preserving way, called Google-Apple Exposure Notification system (GAEN). When two smartphones running GAEN come within range of each other, they exchange information that allows a person, who later tests positive for COVID-19, to anonymously notify those who might have been exposed. MITRE developed and operates the national and international GAEN analytics server, Exposure Notification Private Analytics (ENPA). ENPA collects data from GAEN users, who have opted-in to share their data, in a privacy preserving manner, and conducts analytics to provide insights to jurisdictions.

METHODS: We analyzed data from participating jurisdictions (29.1% of U.S. population) during July through April 2022. We conducted descriptive analyses of the following metrics: exposure notifications (ENs) sent to users, user interactions with ENs, user vaccination status, server-key uploads to notify contacts, and encounter characteristics.

RESULTS: A total of n= 2.10 billion device days were analyzed. In total, n=2,762,617 ENs were sent to users, of which 74.1% were viewed, rather than dismissed or ignored. Among the users who tested positive for COVID-19 and shared their vaccine status (n=3,101), 55% reported being vaccinated. Among users with COVID-19 who uploaded a server-key to notify their contacts of possible exposure (n= 227,663), 14.7% had received an EN within the prior 14 days. Among all encounters (n= 2.82 billion), 59.2% were <5 minutes, and 2.2% had attenuations of 0-60dB (proxy for short distance).

CONCLUSION: The use of GAEN and ENPA demonstrate how digital tools can be deployed to effectively support pandemic response. Future analyses should examine the acceptability of these tools among users and cost-benefit to PHAs.