June 18 - GAEN app presentation and discussion
link to recording: https://covidsafepaths.slack.com/files/U010M1QJGMB/F015PPP0H8B/digital_contact_tracing_tech_alliance__2020-06-18_at_15_04_gmt-7_
Welcome to Emperatriz Garcia
Welcome to Dr. Thomas Kingsley and Walter Cramers from the Mayo Clinics
Presentation of the GAEN API and the PathCheck GAEN app
Sam Zimmerman:
GAEN (Google Apple Exposure Notification) is an API built by Google and Apple. It is built for Health Authorities and only one app will be allowed (“entitled”) to use it per jurisdiction. In the months to come, it will become an OS level feature enriching users' contribution to the network.
GAEN is proximity based, not location based. It communicates only that you were near someone else, using Bluetooth. It generates no PHI, no GPS location data, and is deeply decentralized (Apple / Google have low trust with HAs protecting that information)
Bob Mallon:
The PathCheck GAEN app notifies you when you were “exposed” and for how long using Bluetooth. Users can only run one app (GAEN) at a time, but can download multiple apps (tourism, travel). The app will continue to use Bluetooth independently of other Bluetooth settings on the phone. There is some variability with BT distance measurement.
Exposure notifications are given with date and duration, but not time of day to help protect the identity of the exposure. Following the notification of exposure, there is a call to action in the form of a self-assessment and information download from the (active) registered HA (including using a web view of the HA info site.)
If you have a positive diagnosis, you push your info out to the GAEN server and others you have been in contact with are notified anonymously.
Questions
Vince Munoz: Have you conducted any ground truthing with device to device connection, environmental conditions, OS to Android, etc.
Sam: there is some research on signal attenuation, some public research on Bluetooth blind spots generally and we will be looking further into this in the coming weeks.
research paper on Bluetooth signal strength and contact tracing: https://covidsafepaths.slack.com/files/U011J24HM3M/F016242GYUR/bluetooth_rssi_study.pdf
Dr. Thomas Kingsley: The Bluetooth method doesn’t address other risk factors like if a mask is being worn, is the contact indoor or outdoor, etc. Ultimately a research study is going to have looking into how risky the behaviors, how effective is the app going to report that.
[unknown]: To assess risk from phone data one would need to consider exposure to people who were not infected as a baseline.
Sam: we are going to expand the app to collect and accommodate the needs around making more useful data. Part of the roadmap, part of our intent.